A basic assumption, implicit in all of the Cost Panel deliberations and recommendations, is that a significant part of the total credibility problem is the declining confidence in the Navy’s ability to estimate, manage, and report costs… One cannot factually demonstrate that the Navy and the nation have fewer ships or aircraft because of this alleged cost credibility gap, but it must be assumed implicitly that the seeds of such a negative outcome have been planted…
It [the cost credibility gap] has resulted in a proliferation of reviews and of layers of management dedicated primarily to the review and checking of the actions of the lower levels of management because the upper levels of management do not have the requisite trust and confidence in the management capability of the lower levels. Overruns occur, management layers are added, new checkers are checking the old checkers, funding fences are developed, and the downward spiral in management efficiency and effectiveness accelerates.
That was from the Cost Panel section of the Jan. 1975 “Report of the Navy Marine Corps Acquisition Review Committee.” Volume 1, Office of the Secretary of the Navy, pp. VII-1.
I think there’s an element of truth to that story. The failure to predict the cost and schedule of an approved requirement makes the military bureaucracy look either incompetent or, what is perhaps worse, deceptive through purposeful under-estimation. OSD, OMB, GAO, Congress, and others thus call for more controls and layers of approval in an effort to create more sound baselines in the future. Zero cost growth is the acquisition goal.
I don’t think the characterization is entirely right, however. Technology development is hard, and the pre-contract process of an already layered bureaucracy prematurely fixes the design despite a lack of empirical evidence. And so the requirements likely to be approved by political leaders are probably not the ones that would have been selected with hindsight. Meeting an arbitrary cost or schedule target doesn’t really matter if the requirements aren’t the right ones, and the controls throughout the process hinder the adaptive learning process.
So the lesson isn’t to make “better” estimates of the cost of a ten-year long development program. It is to proceed incrementally, and to create options along the way which can be used opportunistically. Cost and schedule vs. performance is certainly an important way to think about value creation, but it would seem to only make sense in “chunks” that are coherent to decision makers and not too large or too far into the future.
Also, presuming “zero cost growth” equates to good program outcomes is not correct, due in part to whether the approved costs were justified with inflated figures or whether the requirements created unintended consequences.
Leave a Reply