Keith Cowing
Friday, October 10, 2008
Last year, before Ed Weiler came back to NASA Headquarters to be the Associate Administrator for NASA's Science Mission Directorate, an internal cost study was done to see how much SMD overruns every year on its various projects and missions. That study showed $5.4 billion of cost increases over a 4 year period. The goal posts for this study involved measuring costs starting from either the cost contained in the FY 2005 budget or, if the project started later, from its Phase B cost. The costing period then extended forward to the amount that it had increased to for the FY 2009 budget submission.
This large lump of overruns did not sit well with Ed Weiler, so he decided to order folks to make it go away. How to do it? Simple: just move the goal posts. In so doing, you use Congressional guidance (Nunn-McCurdy) as a smoke screen to hide the true magnitude of cost increases.
According to sources inside SMD at NASA Headquarters, Weiler's approach was to forget the numbers that NASA and the scientific community originally bought into when missions were agreed to. Instead, Weiler directed that SMD now use the numbers that arose down the road - after a mission had reached Phases C/D - and then to look at cost growth from that point forward to FY 2009 budget. In other words, Weiler decided to pick numbers that were "more mature" as a starting point.
Of course, this approach resulted in a smaller number since the growth from buy-in to Phase C/D was "forgotten". Weiler also ignored additional costs that were incurred to enhance scientific return and other factors deemed beyond the scope of NASA's responsibility. The new number? Only $1.5 billion in cost increases over 4 years as opposed to the earlier $5.4 billion figure.
Sounds much better that way, right? Alas, this is simply another example whereby NASA cooked the books to make a bad situation look less bad.
Of course, in his defense, Weiler will say that he used metrics and thresholds specified by Nunn-McCurdy. Congress may direct NASA to look at numbers and cost growth that way, but this approach simply does not cover the actual cost growth that has occurred from the original number NASA bought into to start with.
One example is the game being played with real cost of the Mars Science Laboratory. MSL's cost has gone from an inital $650 million (recommended in the Decadal Survey as a "medium cost category" mission) up to the current estimate of over $2 billion. If you use the Nunn-McCurdy goal posts you do not start to track cost increases at the initial $650 million buy-in figure, but rather you start counting at the level that $650 million grew to i.e. $1.4 billion.
When you move the goal posts to this new point, $750 million in cost growth just disappears. This way, NASA can cite Nunn-Mcurdy and say "it is a 30% increase from $1.4 billion up to $2 billion". NASA can now ignore the cost growth from $650 million up to $1.4 billion as if it never happened.
But things have been getting so bad on MSL that even this approach cannot completely hide a serious (and growing) cost problem - one that threatens the vitality of NASA's entire planetary exploration program.
In the end, the process of sweeping these cost increases under the rug means that the taxpayers get cheated out of the full story as to how the cost grew on this mission. The truth is obvious: NASA does not want them to know.
Editor's update: in a telecon with reporters today, a telecon that was supposed to discuss “technical and budget issues” on MSL, NASA personnel more or less avoided providing any specific budget news and tried to shift the discussion back to technical issues.
Ed Weiler stated that his team had a third meeting with NASA Administrator Griffin to discuss MSL. We said that NASA has "made significant technical progress since the May meeting" and "we are heading toward a March 2009 launch".
When asked to come up with a cost impact for current issues, Weiler said "we do not have exact numbers" and that NASA has "no exact estimate from JPL. We are going our own analysis." When pressed for exact numbers, Weiler said "I am not at liberty to pass out numbers."
The MSL cost increase was described as being from an estimate of $1.6 billion in August 2006 to an estimated $1.9 billion today.
When asked if funds will be needed and where they would come form, Weiler said "When we know the final cost in 2009 we'll first look within the Mars program and then outside the program." He noted that there are some cost phasing techniques that can allow resources to be freed up.
When asked again to described what the final cost for MSL would be, Weiler said that "those numbers are being developed. We'll work with OMB and Capitol Hill. It is clear that funding is needed if we go in 2009."
When asked to comment on why an initial MSL cost number of $1.6 billion was used as a basis to calculate cost overruns and not the original $650 million buy in figure, Weiler said "the way NASA accounts to Congress - the cost that NASA commits to - is the cost that NASA buys into - is at Phase C. You do not understand the cost of a mission until Phase C and that is what we have to report to Congress and that is What Doug is correctly quoting."
P.S. If you want to see another example of where NASA constantly changes the cost of a mission, but never admits the true cost, have a look at these posts regarding the Mars Phoenix Mission. Not only did the cost change virtually every time NASA talked about it, they only admitted $100 million that had not been included this year when I pushed the issue. NASA then continued to use an old and inaccurate cost whenever they talked about the mission. Why should anyone believe any cost numbers coming out of NASA?
Why Does The Official Cost of Mars Phoenix Keep Changing?
The Actual Cost of Mars Phoenix is $520 Million
