Today’s New York Times features an op-ed by former NASA associate administrator Alan Stern on NASA’s cost overrun problems. Using Mars Science Laboratory (at least $2 billion now, triple original cost projections) and the James Webb Space Telescope (about a fivefold increase over original projections of around $1 billion) as examples, Stern argues that NASA and its Congressional overseers need to pay much closer attention to costs, particularly in the early stages of a mission, to prevent continued cost growth from eating away at agency programs. “The new presidential administration could begin by accounting for cost increases more honestly, using the initial basis on which missions are started, rather than today’s practice of neglecting certain kinds of cost escalation,” he writes. “And NASA should be charged to reduce or cancel development projects that are not performing to cost. Of equal importance, Congress should turn from the self-serving protection of local NASA jobs to an ethic of responsible government that delivers results.”
In today’s issue of The Space Review, Stern offers a more general prescription for the space agency: make itself more relevant to the public “by combining NASA’s space exploration portfolio with new and innovative initiatives that address hazards to society, make new applications of space, and foster new industries.” These efforts would include new programs in the Earth sciences and aeronautics as well as stimulating commercial human spaceflight. While vague on the details of how this would be done, he notes that this could be accomplished with current NASA budgets—thanks, not surprisingly, to “an aggressive effort to generate new resources from cost control”.
Dr. Stern makes some good points, but no one can deny that he was also a big champion of Griffin’s implementation of VSE. In this view, NASA space science is clearly subordinate to development of human spaceflight systems, most notably the Ares family of launchers. The delays and bleak technical prospects of the Ares program make the difficulties with MSL, JWST and other science missions pale in comparison.
NASA needs more than just a dose of accounting prudence. It’s overall mission is languishing under a myopic, rigidly imposed focus of sending people to the surface of the Moon, while detracting from its broad purpose of exploring space on multiple fronts.
By changing a few nouns and acronyms, Stern’s NYT op-ed would apply equally well to the NRO. As, of course, people have been pointing out since the 1990s.
Cancel msl and there’s nothing going on with the us mars effort until the end of the decade or beyond. It would take a serious idiot or a someone with a hidden agenda to suggest that. I don’t think Stern is an idiot or I suspect the latter. What pet project he likes didn’t make the funding cut?
As for Constellation, it needs a serious review.
michelle: …It would take a serious idiot or a someone with a hidden agenda to suggest that. I don’t think Stern is an idiot or I suspect the latter.
I agree that it is the latter, and is most likely aimed at his positioning to be the next NASA Administrator. However, budget discipline, whether Agencies abide by it or not, is needed across the board. It is a problem that NASA has had to grapple with for decades.
I suspect that Dr. Stern’s past enthusiasm for Constellation will diminish if he becomes administrator. He seems much more genuinely enthusiastic about the potential for the private sector than Mike Griffin has been.
Stern did fight the good fight while at NASA, but this op-ed misses several things and distorts others.
He mentioned one of the key reasons why missions go over budget almost in passing: “The problem continues with managers and contractors who accept or encourage such assignments, expecting to eventually be bailed out.”
The reality is that many missions that “go over budget” are underbid in the first place. People–and this includes contractors as well as NASA managers–propose missions at a lower cost knowing full well that they will grow, but they will not be killed.
The James Webb Space Telescope, which he cites, is the best example: nobody who was involved in that mission at its start honestly believed that it was a $1 billion mission, not the astronomy community, the NASA managers, or the contractors. They all knew that it was around a $4 billion mission (overall lifetime). But they all lied to themselves and each other, got the mission started, and then had to find the money. They did this by wrecking the rest of the astronomy program (i.e. delaying all the other missions). The reality is that once this was accepted, JWST actually became a well-run program and it has hit all of its cost targets for the past two years. JWST is no longer a mess (although time will tell), but it was sold under false pretenses. Stern ignores this factor in his op-ed.
Stern also gets the Mars Science Laboratory issue slightly wrong on two counts. First, it is not as over-budget as he claims. He has used the wrong starting point for the mission cost. He is correct that MSL has suffered some crazy cost-overruns. Some of these are totally understandable, the result of the engineering challenges being more difficult than initially expected. Some of them are not understandable and it has been clear for at least a year now that JPL does a bad job of estimating their projects and even close oversight from Headquarters has been unable to correct this. My own solution to this is to introduce more competition into the system, which is something that Dan Goldin started to do in the mid-1990s by encouraging the Applied Physics Laboratory to get involved in planetary missions. If JPL starts to believe that they can actually lose missions to competitors, then they may get their act together. So build APL up into a competing center of excellence on planetary missions.
The second thing Stern is wrong about is the reason why NASA is not canceling MSL. He claims that it is because of “sunk costs.” The reality is that MSL is a good mission and has universal support in the planetary community and _also_ that without MSL, NASA’s Mars program will fall apart for the next decade. The community agrees that MSL is the proper logical step in the Mars program. The MER rovers are good, but very limited in where they can land, how far they can go (five years on Mars and they have traveled 12 km), what terrain they can cover, and what they can do. The interesting spots on Mars are in higher latitudes and altitudes and rougher terrain. Plus, you have to bring more instruments to bear than the simple MER instrument suite. MSL is a good, well-vetted solution. MSL also makes sense from an engineering perspective as well: if your goal is ultimately Mars Sample Return (which was one of Stern’s goals when he was at NASA), then you need better landing technology than airbags. MSL was designed from the start to develop landing techniques that can later be applied to Mars Sample Return. Cancel MSL and MSR is not doable unless you find another technology development path.
MSL is the right project at the right time. The cost overruns are bad, but the solution is not to kill MSL, but to find another way to impose rigorous oversight and perhaps competition to JPL’s programs. And maybe find a good way of punishing them for not controlling costs, like killing one of JPL’s other missions.
“In this view, NASA space science is clearly subordinate to development of human spaceflight systems”
This has always been true at NASA. Space science has been approximately 20% of the overall NASA budget for decades, and only started increasing under Goldin. If you hate Constellation, fine, but you cannot blame the AA for science for a situation that has always been true.
Alan’s suggestion is that NASA SMD ought to manage costs better is kind of astonishing. Jeez, as if NASA SMD hasn’t been struggling with that for the last decade or two. It’s a valid point, but it sure isn’t a novel one. If NASA wants to manage costs better, it’s going to have to make some additional up-front investment in project cost vetting, and raising TRLs on technology. It’s not going to happen by saying it ought to.
His point about NASA having to be more meaningful to the public is dead on. This is not about being meaningful by offering to put outposts on the Moon. It’s about offering some national pride in achieving widely agreed upon national objectives. An outpost on the Moon is not a widely agreed upon national objective, though some have the strange illusion that it must be. It’s certainly far for obvious that humans going to Mars is a widely agreed upon national objective, but that’s so far off it’s not worth worrying about.
Regarding uncrewed deep space exploration competition, isn’t that what we used to have with Ames managing Pioneer 10 & 11, Langley managing Lunar Orbiter & Viking, and Goddard running other non-planetary missions? I’d have to dig pretty deep to track down the details (and to double-check my memory), but when did the notion of NASA center competition like this go away? Who decided that only JPL can manage our deep space remote-controlled spacecraft?
“I’d have to dig pretty deep to track down the details (and to double-check my memory), but when did the notion of NASA center competition like this go away? Who decided that only JPL can manage our deep space remote-controlled spacecraft?”
Your memory is false. APL is managing Stern’s New Horizons Pluto mission, as well as the MESSENGER mission to Mercury (depends on your definition of “deep space”).
There is some competition today. My point was that it needs to be increased. At the moment, there is a general view that only JPL can manage the very large “flagship” missions and only JPL can do Mars landings. It is true that they have the most expertise, but it might help to threaten that.
I see a bit of inter-center tension here. Stern is now working for Odyssey Moon, which has partnered with Ames to launch a spacecraft to the moon. There’s a definite feeling at Ames that JPL is out of control and tends to find exceedingly expensive ways to do missions. A sense that Ames could do it cheaper. Stern seems to be challenging some of that.
Question: Does Stern mention anything in these pieces about the conflict of interest that delayed one of the future Mars missions? That seems to have done damage to the program. Yet, to my knowledge, the conflict has never been publicly explained to the people who are paying for the mission.
“If NASA wants to manage costs better, it’s going to have to make some additional up-front investment in project cost vetting, and raising TRLs on technology. It’s not going to happen by saying it ought to.”
That’s not entirely fair to Stern. He sought to do a number of new things to impose cost discipline. He said on numerous occasions that one of the things he learned from his predecessor (Mary Cleave) was that he could not kill any program, not even a Discovery class mission (like Dawn) that had gone over its cost caps. He also realized that he could not fire a principal investigator. But he realized that he could fire project managers and NASA personnel involved in a mission, and sought to do that to put people on notice that overruns would not be tolerated. Kepler is an example where he made a lot of noise and the project managers got their act together.
Another thing that he sought to do was to increase the requirements of experience for principal investigators. His argument was that the first large project that somebody manages should not be a $400 million government program. That was an unpopular decision. Those requirements were eliminated once he left. Another thing he sought to do was to increase training opportunities for people who would eventually become managers. That was a more popular decision, but also seems to have been abandoned.
In order to keep programs in their cost box, you need a combination of carrots, sticks, and training. The upfront studies, cost vetting, and technology development is important, but is not sufficient to do this.
“I see a bit of inter-center tension here… A sense that Ames could do it cheaper. Stern seems to be challenging some of that.”
It’s doubtful that this has prompted him. His current statements are entirely consistent with what he did as AA, so you cannot say that this is inspired by Ames.
Ames can do some things cheaper, and they are trying to get into the business of building small, faster spacecraft. However, Ames cannot really compete with JPL. They don’t have the people or resources. Ames still remains very much an aeronautics center with an excellent but small space science team. They don’t have the ability to handle larger flight missions.
“Does Stern mention anything in these pieces about the conflict of interest that delayed one of the future Mars missions? That seems to have done damage to the program. Yet, to my knowledge, the conflict has never been publicly explained to the people who are paying for the mission.”
No, and he cannot for legal reasons.
I have no direct knowledge of what happened, and not even good indirect knowledge. NASA did a good job of not leaking on this subject. However, it is pretty clear that what happened in that case was not NASA’s fault and the agency did the only thing it could do, although it cost them tens of millions of dollars. At some point late in the selection process, an individual on the review team–not a NASA employee–accepted a job offer from one of the two institutions involved in the competition. This does not even have to be a job offer involving anything to do with the spacecraft; the person could have simply accepted a job at the same university but in an entirely unrelated department. Once that happened, it opened up NASA to a legal challenge from the loser. A legal challenge would have delayed the award, caused the spacecraft to miss the launch window, cost a lot of money and opened NASA to criticism.
There are several recent examples where government contract awards have been overturned. The most obvious one is the USAF tanker deal, which happened after the NASA event, but there were several big ones that predated this. NASA had plenty of examples of how things can go bad. My own guess is that they have a directive from OMB that they are to avoid this at all costs. So the agency’s hands were probably tied at that point and they had to follow OMB rules (and they probably realized this would be the right thing to do no matter what).
The only thing NASA did wrong in all of this was the way that they released the information, sneaking it out right before a holiday. That was sloppy. But everything else seems to indicate that they were a victim of some individual and this was a case where the government did the best that it could do under the circumstances.
MSL has not tripled in cost. The first NASA cost estimate for the mission was $1.4 billion, not $650 million. See Jim Garvin’s letter to the editors of Science and Space News and Alan Stern’s response here (add http://www.):
nasawatch.com/archives/2008/11/msl_cost_overru.html#more
People living in glass houses should not throw stones. MSL has overrun by 26%, but Alan Stern’s own New Horizons mission to Pluto has overrun by 35% so far, going from a $500 million cost estimate at selection to $675 million a month before launch. See (add http://www.):
space.com/spacenews/archive05/Horizon_121205.html
Charity starts at home. If Dr. Stern is really concerned about the impact of mission cost growth on NASA’s planetary science portfolio (and not just polishing his record as Associate Administrator for Science or running for NASA Administrator), then he should offer up an op-ed admitting to the greater percentage cost growth on the mission for which he was principal investigator and analyze the reasons for that cost growth so that lessons can be learned for the future.
From a taxpayer’s perspective, there should be no cost growth on any NASA mission, and legitimate criticism of truly outrageous cost growth (as on JWST or NPOESS or any number of military space programs) is certainly warranted. But hypocritical and potentially self-interested mud-slinging that blows a fairly typical cost increase way out of proportion does little to inform or benefit future programs.
By Dr. Stern’s logic, most NASA missions, including his own New Horizons mission, would have been terminated before flight for small to modest cost increases late in development. Maybe that’s what he would try to do were he back in charge, but it’s hardly good program management or a wise use of taxpayer resources.
FWIW…
ANON.SPACE: If Dr. Stern is really concerned about the impact of mission cost growth on NASA’s planetary science portfolio (and not just polishing his record as Associate Administrator for Science or running for NASA Administrator), then he should offer up an op-ed admitting to the greater percentage cost growth on the mission for which he was principal investigator and analyze the reasons for that cost growth so that lessons can be learned for the future.
Anon,
I agree that it is fair to ask Dr. Stern to publicly explain the reasons for the 26% growth in New Horizons, and the lessons learned. Although I doubt it was your intent, I am concerned that your message comes off as trying to shut Dr. Stern up. (“Hey, if you criticize my program, I will criticize yours” can be a destructive tactic that discourages free, open, and rational policy discussions.)
I do believe that this issue is more complex that you make it out to be.
1) NASA leadership (or somebody) pulled a bait & switch — The MSL was sold to the NRC on the basis of being a $650 million mission. It was scored a high priority based on an estimated cost $650 million. Somewhere, somebody turned this into a $1.4 Billion mission.
That is a pretty important change (an over 100% cost growth), which was made behind closed doors, and which had a significant impact.
Somebody needs to do some explaining.
I also think it reasonable to consider — as a matter of policy — that if we truly want to empower the NRC prioritization process, that we should send consider a new policy of automatically sending huge program changes back to the NRC for review/comment before approving the major change (and all its impacts).
2) Some sources are now suggesting that MSL will cost as much as $2.1 Billion. If true, the cost growth of MSL from even the higher $1.4 Billion cost number will be 50%, not 26%.
FWIW,
– Al
“I agree that it is fair to ask Dr. Stern to publicly explain the reasons for the 26% growth in New Horizons”
Just to keep the record straight, it’s New Horizons that has experienced 35% cost growth to date (or at least to launch per the Space News article I linked to above). It’s MSL that’s had 26% cost growth to date (per Garvin’s letters to the editors of Science and Space News linked to above).
“am concerned that your message comes off as trying to shut Dr. Stern up.”
I’m just pointing out the hypocrisy and misleading statements. For a PI to take another mission to task for a 26% overrun to date, when the mission that PI is in charge of has experienced a 35% overrun, is very hypocritical. And for a former NASA Science AA to treat a pre-pre-Phase A cost estimate in an NRC report as a hard cost constraint is misleading and inappropriate in the extreme. A former NASA Science AA should know better.
As I said in the last paragraph of my prior post, there’s nothing wrong with being critical of NASA cost overruns, even small ones. The taxpayer deserves that. But such criticism should take the high ground by admitting one’s own culpability for past overruns, in addition to pointing out the failings of others. And it should be done without misdirection and misleading statements.
There’s nothing I can do to stop anyone from making any statement they wish to make. But if I know (and more importantly can back up) that a statement is hypocritical or misleading, then I (or anyone else) should arguably point such out.
“That is a pretty important change (an over 100% cost growth), which was made behind closed doors, and which had a significant impact.
Somebody needs to do some explaining.”
It’s hard to see that much of anything was done behind closed doors. Very early cost estimates are printed in the NRC surveys. And mature life-cycle cost estimates for the same missions appear in NASA’s annual budget request to Congress. It takes a few minutes to go looking in a couple documents to compare the numbers, but they’re there in black-and-white for anyone who cares to see.
“I also think it reasonable to consider — as a matter of policy — that if we truly want to empower the NRC prioritization process, that we should send consider a new policy of automatically sending huge program changes back to the NRC for review/comment before approving the major change (and all its impacts).”
Agreed. When a mission goes from being a multi-hundred million dollar class mission to a multi-billion dollar class mission, the justification for the mission — and how that justification weighs against the justifications for the other missions that will be deferred/terminated to pay for the cost increase — needs to be reassessed by the relevant science community at large. Just by the uncertain nature of the very early cost estimates that must by necessity be used in NRC surveys, this is going to happen, and there should be a mechanism to deal with it (which there isn’t now).
But we (Stern included) shouldn’t pretend that those very early and very uncertain cost estimates have the same standing as, say, a cost estimate at SRR or PDR. It’s comparing apples and oranges. What’s happened on JWST (or NPOESS or various milspace programs), for example, is in a totally different ballpark than what’s happened on MSL (or New Horizons).
“Some sources are now suggesting that MSL will cost as much as $2.1 Billion. If true, the cost growth of MSL from even the higher $1.4 Billion cost number will be 50%, not 26%.”
If that happens, you’re right — there’s nothing I can argue with in those numbers. But if it does happen, it’s still not in the same ballpark as a multi-100% JWST (or NPOESS or milspace) overrun. And that’s one of several places where Stern was being misleading.
FWIW…
“the very early cost estimates that must by necessity be used in NRC surveys, this is going to happen, and there should be a mechanism to deal with it (which there isn’t now).”
Doing it via the NRC route is not very practical. For starters, the decadal surveys are large affairs, involving dozens of committee members, many meetings, and lots of community input. If you are going to return to the NRC to ask for a reconsideration of priorities, then is it fair to get only a small subset of the community to consider something that was previously decided based upon a larger community effort? In addition, the NRC is not a fast process: you have to create a committee (many months), have them hold meetings, write a report, and then send it through review (review alone is about two months). All of this is to ensure objectivity and independence, so if you start cutting into any of those parts, you will lose some of that. The NRC is not the proper method for providing such input. In the past, the NASA Advisory Council and its subgroups provided a method for doing this, but that all changed.
“New Horizons mission to Pluto has overrun by 35% so far, going from a $500 million cost estimate at selection to $675 million a month before launch.”
I don’t believe that this is an entirely accurate claim. New Horizons was the first of the New Frontiers class medium size missions (and was actually grandfathered in). New Frontiers missions were budgeted in the decadal survey (the same one Stern quotes for MSL) at up to $650 million. I’m not sure that the Space News article is accurate that New Horizons was originally estimated at $500 million.
In fact, while throwing around claims such as this, we’re missing some important details. One of them is that significant costs were actually _imposed_ upon New Horizons by NASA. The agency forced several requirements on the mission that drove up costs. I have a list of them somewhere, but one of them was the requirement that NH be designed for either Atlas or Delta launch until late in the design process, which placed substantial burdens on the design team (tens of millions). I also believe that the cost of the RTG was significantly higher than what the project was originally told. NASA admits this. Once you factor out these NASA liens on the project, the actual New Horizons overrun is substantially less. (This is not an unusual situation either. Juno suffered a similar fate when NASA selected it and then immediately delayed the launch by two years, which increased the cost substantially–because the agency has to pay for the standing army of workers and cannot simply tell them to come back in two years when NASA has the money.)
My point is that beyond the issue of costs, there are many other important variables. We have to ask _why_ some of those overruns occurred. Were they foreseeable? Were they the result of bad choices? Or were they the fault of someone else?
A thought just occurred to me based upon the Space News article that cites a $500 million cost for New Horizons. The article states:
“NASA held an open competition to find a cheaper way to do a Pluto flyby and ultimately selected the $500 million New Horizons proposal but did not request any funding for the mission. Congress funded the New Horizons program as a budget earmark for two years before NASA finally got the message and started requesting money for the mission.”
If so, then this can explain at least part of the increase. NASA “selected” the mission at $500 million in 2001, but did not actually start funding it until 2003. That alone would have driven up the cost of the mission, at no fault to the PI. After all, if you find a car you like at a dealership, but wait two years to buy it, inflation will have driven up the cost.
“After all, if you find a car you like at a dealership, but wait two years to buy it, inflation will have driven up the cost.”
Actually if you saw a 2008 model car, and waited two years to buy it, then the price would have depreciated in value.
A better example would be, you found a set of house plans, got an estimate for how much it would cost, then waited two years to have it built, you would then may see the construction costs rise.