Costs on the Books

Avatar

Patrick

Patrick is a mid-40 year old geek with an undergraduate degree in mathematics and a master's degree in Information Systems. Nothing he says here has anything to do with the official position of his employer or any other institution.

Related Post Roulette

33 Responses

  1. Avatar wardsmith
    Ignored
    says:

    Great OP Patrick, looks like you fixed your writer’s block. 😉

    I too have friends who worked for JPL and NASA. To your point about 22 companies spending more than one NASA, completely agreed. However there’s another element going on here. As a monolithic enterprise NASA has a tendency to prune promising lines of research and inquiry dur to bureaucratic and political infighting My father saw the same thing happen constantly at AEC. It all comes down to who gets money budgeted to their cause and why.

    Also, and let’s be really clear here, NASA doesn’t really “build” anything. They manage contracts for other companies who as you said, bear the burdens of researching, testing, producing, delivering and so on. Boeing can do some amazing things on the space side because of the cash cow business that is their commercial aircraft division or their military group. NASA handing the same contract to a startup would entail said startup wasting (there’s the word again) a lot of money building what is to Boeing redundant infrastructure. Of the $500M Solyndra blew, how much was wasted on the fancy building, fancy offices and fancy furniture?Report

    • Avatar Jesse Ewiak in reply to wardsmith
      Ignored
      says:

      It’s a good thing there’s no such thing as political infighting and overwhelming bureaucracy in the private sector. Now, off to try to get a hold of anyone at AT&T with working brain cells.Report

    • Avatar Patrick Cahalan in reply to wardsmith
      Ignored
      says:

      As a monolithic enterprise NASA has a tendency to prune promising lines of research and inquiry due to bureaucratic and political infighting

      Oh, that’s totally true, Ward.  This is true of all sorts of large organizations, and NASA certainly isn’t immune (hence the “NASA has cultural problems. NASA has efficiency problems. As an organization, there’s a lot that could be done to make NASA a whole lot better at being NASA than they are.”)

      This is another benefit of smaller and/or shorter lived organizations: they don’t have as much embedded culture.  Less baggage when it comes to decision making.

      In certain types of emergent problem spaces, smaller more agile companies or organizations will be able to work much more efficiently because they’re more willing to jettison assumptions than large organizations, and in emergent problem spaces, assumptions can be big problems.

      In more established, well-known problem spaces, larger organizations can outperform smaller organizations because they’re more likely to have the institutional knowledge that enables them to tackle the well-known problem most efficiently.

      Aligning the two is what gets you real efficiency.Report

  2. Avatar George T
    Ignored
    says:

    The post is replying to an incorrect claim, that Space X spent half as much as NASA would to develop the Falcon 9.

    Here is NASA’s report on the cost difference.

    Results: “The activity estimated Falcon 9 would cost $3.977B based on NASA environment/culture. NAFCOM predicted $1.695B when all technical inputs were adjusted to a more commercial development approach.

    Space X spent $390 million for the Falcon 9, a number which includes the $90 million spent developing the Falcon 1.

    The difference is a factor of ten.Report

    • Avatar Patrick Cahalan in reply to George T
      Ignored
      says:

      Thanks, George, I’ll read that and digest.

      Still, the takeaway from the post is supposed to be that the correct comparison isn’t between the successful candidate vs. the government entity.  The correct comparison is comparing all of the candidates vs. the government entity.

      NASA can undoubtedly learn a lot from Space X, there is no question there.Report

      • Avatar James Hanley in reply to Patrick Cahalan
        Ignored
        says:

        the takeaway from the post is supposed to be that the correct comparison isn’t between the successful candidate vs. the government entity.  The correct comparison is comparing all of the candidates vs. the government entity.

        Excellent takeaway message, too.  Nice post, PC.Report

    • Avatar DensityDuck in reply to George T
      Ignored
      says:

      Nope.  The “factor of ten” was based on an inaccurate estimate of SpaceX’s costs; when the estimate was revised, the cost estimates dropped sharply.  The new estimates, when calculated on the same cost basis, either “firm fixed price” or “cost plus fee”, come in with SpaceX about half of what NASA would expect to pay to produce the rockets.Report

  3. Avatar Tod Kelly
    Ignored
    says:

    Nothing to add, other than this was an excellent post.Report

  4. Avatar Will H.
    Ignored
    says:

    I’ve done work at Kennedy Space Center (it’s within the jurisdiction of my home local).

    I don’t think these private companies had armed guards at every gate.

    In those instances where I was at a facility where industrial espionage was a huge concern, it always seems to be a chemical plant.  That’s just my experience.

    I don’t know if I’m supposed to say this or not, but NASA really does have some obsolete equipment out at Kennedy.Report

  5. Avatar Jason Kuznicki
    Ignored
    says:

    Suddenly, that sounds like kind of a crappy economic trade-off, don’t it?

    The first time, yes. But you’re making a very serious economic mistake to only look at costs the first time around.

    As with so many other activities, there are diminishing marginal costs.   In particular, there is no additional cost for research.  So the private sector still definitely wins.Report

    • Avatar Patrick Cahalan in reply to Jason Kuznicki
      Ignored
      says:

      This is a really good point, Jason.

      I’m not so sure this automatically leads to “the private sector still definitely wins” in all cases, though.

      If all you want is a rocket built to spec… and then N more rockets, yes.  But at least in the case of NASA’s overall mission, a very large portion of their work isn’t susceptible to reuse in that way.  For the Saturn rocket & the Space Shuttle (just to give two examples), probably and likely.  For the Mars Rover, not so much; most of what is learned building the Mars Rover isn’t going to be of as much assistance building a Titan rover.Report

  6. Avatar DensityDuck
    Ignored
    says:

    A:  SpaceX wasn’t involved in the X-Prize.

    B:  Let’s put some numbers in this.  If you examine the two most representative SSTO-technology development efforts (DC-X and X-33) you find that DC-X spent about $70 million, and X-33 spent over a billion.  Comparatively, Scaled Composites (and Paul Allen) put about $25 million into Space Ship One. 

    C:  The situation of the X-Prize–an award offered to the winner of an open competition–is pretty unique in the aerospace field. The correct comparison may be “all of the competitors vs. a government entity” but very seldom does anyone build the proposed system and have a fly-off.  More commonly it’s Dueling Powerpoints and the prettiest one gets picked. 

    You say that we should compare the totality of entries to the government, but in the scenario I suggested–SpaceX versus NASA–there’s only one entry.  And even at that, it’s worth asking what makes any one cost estimate be so much lower than the government’s.  What is that company doing that makes them so much less expensive?  Or, perhaps, what’s the government doing that makes it cost so much?Report

    • Avatar Patrick Cahalan in reply to DensityDuck
      Ignored
      says:

      A: Re-reading my post, I could see that someone might be confused.  I’ll clean it up tomorrow.

      More commonly it’s Dueling Powerpoints and the prettiest one gets picked.

      Yes.  That’s a bad way to do it.

      You say that we should compare the totality of entries to the government, but in the scenario I suggested–SpaceX versus NASA–there’s only one entry. 

      But that’s part of the point, Duck. If you compare NASA to just one company that succeeds at a part of the mission that NASA does, the company is going to win every time because they’ve got a smaller mission and you’re only looking at the winner.  When you have publicly-financed monolithic R&D, the public entity is carrying all the loser branches of research on the books.

      The Falcon 1 can carry 420kg into LEO, but that’s it; it can’t get to GTO.  An Atlas III can carry twenty times that amount to LEO, and ten times as much to geosynchronous orbit.  SpaceX has a singular mission: get small payloads into low-earth orbit.  They’re focused on it.  If they have a design consideration which could go a cheaper route focused on that single mission, or a polymorphic, more expensive route that could wind up working in both an LEO rocket and a GTO one, they have an incentive to take the cheap route because they have a less general mission.

      There are a lot of other things going on here, too.  Look at who works at Space X.  Space X went out and hired the cream of the aerospace industry.  Where did all those folk excepting the CEO work before?  Boeing, the Aerospace Corporation, Kennedy.  Building stuff for NASA.  Space X bought all that expertise that was developed during a couple of decades working on government projects.  It’s not like Space X had to create the profession of flight engineer from nothing. If they were a real startup, they’d have had to do a lot more beta testing and blown up a lot more stuff on the pad.

      Here’s another one: I know a guy who was working on the Mars Rover.  At one point, he was working with a Chinese subcontractor for part of the project.  He’s sitting there, fuming, because the Chinese subcontractor has a problem that they can’t figure out, and it’s causing delays and cost overruns.  But he knows what the problem is, and what’s more he knows what the solution is… but the contract that was signed with the subcontractor (at Congress’s demand to cut costs) is subject to severe ITAR restrictions.  He’s got a 200 line script and he could just load it up and run the test on their subsystem and show them exactly what their problem is so that they can fix it… but if he does it’s 20 years in the hoosegow for trafficking in missile secrets.  To re-issue the contract would require jumping through another round of ITAR restrictions.  So he’s sitting on his ass for three weeks until they figure it out.

      That’s not to say that it’s not a damn good idea to work the model.  Anybody that works at NASA will tell you over a beer a long laundry list of things that they’d like to change.  It’s also true that getting the private sector to compete is very likely going to net you a bunch of great tradeoffs.Report

      • Avatar DensityDuck in reply to Patrick Cahalan
        Ignored
        says:

         “The Falcon 1 can carry 420kg into LEO, but that’s it; it can’t get to GTO.  An Atlas III can carry twenty times that amount to LEO, and ten times as much to geosynchronous orbit.”

        What does Falcon 1 have to do with anything?  SpaceX’s EELV competitor is Falcon 9, and its projected performance is entirely in-family with the Atlas V / Delta IV.  And the NAFCOM study was indeed based on Falcon 9.

        “If you compare NASA to just one company that succeeds at a part of the mission that NASA does, the company is going to win every time because they’ve got a smaller mission and you’re only looking at the winner.”

        …except I’m not comparing “all of NASA” to “one tiny part of the private industry”.  I’m comparing a specific mission.  One system.  And this is based on a cost estimate performed by NASA itself, which declared that private industry left to its own devices could accomplish that mission for half the cost NASA would pay to do it the NASA way.

        You’re just making a different version of the “cherry-picking” response to criticism, which is often valid; but in this case the same cherry-picking was done to both sides of the comparison.Report

        • Avatar Patrick Cahalan in reply to DensityDuck
          Ignored
          says:

          I’m comparing a specific mission.  One system.  And this is based on a cost estimate performed by NASA itself, which declared that private industry left to its own devices could accomplish that mission for half the cost NASA would pay to do it the NASA way.

          I guess I’m not quite getting through, but okay.  I’ll try to restate it.

          You have a point, in that this is indicative of problems at NASA (which is something I’ve never disputed). There are problems at NASA, some of which are cultural and some of which are due to the way they *have* to do business, as a government entity.  The first can be corrected, the second is in many cases an issue with the government forcing itself (for various reasons) to do things inefficiently; some of this is due to regulatory zeal, and some of it is due to audit zeal.

          However, what I’m trying to say is that this is still not a legit comparison, because you’ve drilled too far down.

          Let’s say that NASA has six missions; near-Earth orbital deliveries, GTOs, robotic exploration, maintenance of existing equipment, making foo, and making bar.

          NASA is a monolithic organization.  There is a requirement for a piece of technology that is polymorphic: it is a requirement for making foo, near-Earth orbital deliveries, and robotic exploration.

          If we choose a solution to that technical problem that solves all three use cases – call it the Thurman Unit – it costs $M.  If we choose a solution to that technical problem that only solves one – call it the Jolie Unit – it costs $M-N.

          It may very well be that if we separate out one use case, and compare NASA to the private sector where the private sector solution will always choose the technology that is least expensive for its use case, and thus be cheaper.

          But those other two use cases still need to be met (presumably), and NASA will still have to develop the solution that works for those two use cases.  So now we’ve outsourced “making orbital deliveries” to Space X, and our rockets cost half as much… but making foo and bar is suddenly *way* more expensive, because we don’t have technology reuse; instead of having the $M spread across all the missions it supports, we have the $M spread across two missions and we’re also paying SpaceX $M-$N. We still need Thurman Units, but now we’ve engineered a dependency on both Thurman Units and Jolie Units.

          If we can dump the extra missions, then this is a win, obviously.  Taking a long hard look at NASA is probably a really good idea. If we can coordinate with Space X, we can still get benefits of polymorphism… maybe instead of paying twice as much for having NASA build the rocket, or paying half as much to have Space X build the rocket but get no technology reuse, we can pay 4/7ths as much and have Space X build rockets with a Thurman Unit instead of a Jolie Unit and then we can use Thurman Units across all problems where we need one.

          There’s lots of potentially good ideas here.  But focusing too much on one particular mission with one particular competitor gives a distorted view of how much money you can save.  Just ask everybody that outsourced all their IT in 2000-2007.Report

          • Avatar DensityDuck in reply to Patrick Cahalan
            Ignored
            says:

            I’m not drilling down any further than NASA did in the NAFCOM study. 

            You’re certainly right that a man-rated rocket is much more expensive than a payload-only one (the escape system alone costs more than developing a new unmanned booster!) But that’s not what’s happening here. 

            “Let’s say that NASA has six missions; near-Earth orbital deliveries, GTOs, robotic exploration, maintenance of existing equipment, making foo, and making bar.”

            Assuming that none of these involve human spaceflight (keeping in line with the NAFCOM study) then they all get accomplished with the same booster.Report

  7. Avatar Morat20
    Ignored
    says:

    There’s also the point that Space X — and everyone else — built their stuff using expertise, experience, and technology NASA pioneered. Even for purely private space companies, they are using a ton of information that NASA paid to develop.

    Generally for free.Report

    • Avatar DensityDuck in reply to Morat20
      Ignored
      says:

      Yes, and so does everyone else, including NASA.

      Heck, that makes NASA look even worse–if it costs NASA twice as much to bolt existing technologies together then what the hell are they doing over there?Report

      • Avatar Kim in reply to DensityDuck
        Ignored
        says:

        Making sure people don’t die. And, occasionally, paying people to insert typos into code. Because NASA hires the fucking best people, and milspec is milspec — you can’t use it until it’s been “debugged.”Report

      • Avatar Morat20 in reply to DensityDuck
        Ignored
        says:

        They’re not bolting existing technologies together. They, for the most part, invented those technologies — which Space-X and it’s competitors get to use without paying license fees, patent fees, royalties or anything like that, often employing ex-NASA employees or contractors who are not under non-compete clauses. 🙂

        It then extends those technologies in, yes, a very slow and careful way — because NASA is not allowed, as the original post pointed out, the luxury of going bankrupt or out of business.

        It’s a sole, singular governmental entity, not a froth of creative destruction in which one business succeeds and twenty fail — spending, cumulatively, dozens of times what NASA might. And using, as noted, technology and expertise they didn’t have to pay for developing, but that was handed down from on high.

        I might add that, as my own personal NASA gripe, Space-X does not have Congress constantly dicking with Space-X’s business goals, plans, methodology, and funding.

        You want an engineering challenge? Design a space station on a budget that changes year to year, whose requirements change every other year at the whim of people with no engineering expertise, and when you are required to work with other companies who are randomly assigned — and reassigned — bits and pieces of your every changing project and chunks of your budget.

        Oh yes, and all this done by those “non-experts” who don’t even understand that while it’s costly to design something, it’s considerably MORE costly to tear down and reengineer an existing design, so they expect you to keep your original time-line and budget even though they’ve played games with your budget (increasing it or decreasing it by large percentages year to year) and even with your end goals.

        Frankly, if NASA can make a rocket for only twice what Space-X does under those conditions? They are geniuses.

        Seriously, just check the fun of replacing the Shuttle. How many engineering plans has NASA put forth only to have funding yanked, then reappropriated with NEW goals and design requirements, only to be yanked again?

        NASA could probably save a ton of money if it’s budget and goals were allocated in five year chunks, rather than yearly.Report

        • Avatar DensityDuck in reply to Morat20
          Ignored
          says:

          Congratulations, you’ve answered my original question, which was “how is it that SpaceX can do the same thing that NASA does, using the same inputs, and yet only cost half of what NASA thinks it ought to”.Report

          • Avatar Morat20 in reply to DensityDuck
            Ignored
            says:

            As I said — it’s a miracle it only costs half as much, and that’s with Space-X hiding the true costs of the effot (Space-X plus everyone who failed), given the handicaps underwhich NASA works.

            Hmm. How many companies failed? How much money was lost as sunk costs, money not on Space-X’s books?

            But this particular reality is apparently troubling for you, and you refuse to acknowledge it.

            Space-X got free tech and free expertise — money NASA carried on it’s budgets that Space-X didn’t. NASA has only itself — it carries the budget of every failure along with every success.

            Space-X was one of many companies vying for the prize — but it did not calculate the wasted money on all the failures.

            I bet you I could beat the per-unit price of making practically anything if I was allowed to poach the basic designs, 50 years of knowledge, and a good chunk of the personnel involved.

            Heck, I’d have to be a major screw up to cost more per-unit! After all, someone did half the work for me — free of charge.Report

            • Avatar DensityDuck in reply to Morat20
              Ignored
              says:

              but NASA has access to all of that as well.Report

              • Avatar Patrick Cahalan in reply to DensityDuck
                Ignored
                says:

                Duck, I’m not sure if you’re being deliberately obtuse or you’re honestly not getting the point.  An anechoic chamber is an expensive little facility.  Somewhere between $160 and $250 K to build, depending upon how big you’re going.  A class 10,000 clean room is another pretty expensive capital expense.

                Here’s an example; I don’t know that this is entirely accurate, but it could easily work out this way.  The windtunnel at Ames wasn’t cheap.

                Building these things to enable basic research costs a lot of money, in up-front capital costs.  You need to depreciate those assets over time.  You need to maintain them.

                It used to be the only way you could get access to any of those facilities was to foot the bill for the creation of ’em.  Like NASA did.  Nowadays you can rent them, and the government is predisposed to rent them to you for far less than an equal time charge.

                NASA wants a wind tunnel.  They need one *to exist*, so they build it.  They use it 15% of the time, or whatever.  It cost $27 million to build, and another $N per year to maintain the thing.  But the maintenance cost is almost entirely decoupled from the per-use cost; they can use it for 8 weeks a year and then let it sit and pay the $N in maintenance, or they can use it for 8 weeks a year and then rent it out for a charge of $X per week and pay $N plus $Z in maintenance.  As long as $X is more than the energy cost to fire the thing up – $Z – they’re cutting their costs to rent it out… it would be irresponsible of them *not* to rent the thing out, even if they didn’t get 85% of the maintenance cost back in the rental; even though they’re only using it 15% of the time.

                But when they assign those costs internally, they have to distribute them somehow.  Effectively, they can wind up allocating a very large chunk of the cost of the wind tunnel to their direct or indirect costs for any particular project.

                Depending upon how they do their internal accounting, then… a rocket-building project run by NASA may be assigned the entire depreciation and maintenance cost of that wind tunnel, less the $X nominal charge they collected by letting somebody else use it.

                Effectively, they’re charging themselves quite a bit more than someone else would have to pay to use the same facility. Even if that someone else is making rockets to compete with NASA’s design.

                There are numerous other possible examples.

                NASA isn’t a private actor.  They don’t really have trade secrets to protect (unlike a private sector “competitor”).  They’re not actually in the business of competing with anybody.  The way they account for expenditures is nothing like a for-profit corporation.

                Now, taking a harder approach towards cost-benefit analysis might wind up producing a leaner NASA that can get the job done with less capital costs or whatever.  No argument there.Report

              • Avatar BlaiseP in reply to Patrick Cahalan
                Ignored
                says:

                Heh.  I know a team of software contractors who sat around writing the NASA travel expense application for well over a year.   How very right you are, the way they account for expenditures is, um… (dark hilarity ensues)  nothing like a for-profit corporation.Report

              • Avatar Patrick Cahalan in reply to BlaiseP
                Ignored
                says:

                Don’t get me started on travel charges on the Federal government’s dime.  People around here spend more time dealing with freakin’ travel expenses than you would believe.

                Every U.S. airline can go to hell for their capture of the federal grant process.Report

              • Avatar BlaiseP in reply to Patrick Cahalan
                Ignored
                says:

                I’ve thus far done gigs for, lemme count ’em up,  USDA, US Transportation Command, FBI and US Army.   It’s not the airlines who are screwing it up.   Contracting for the Army, I filed my travel expenses in five minutes every week.   There’s no excuse for NASA’s decrepit and moribund bureaucracy.   The only worse bureaucracy I ever saw professionally was at FBI and I worked on that gig with the same crew who did that NASA gig and they swear it’s worse at NASA.

                Put it this way, I’ve never done less work as a contractor than I did for the gummint.Report

              • Avatar Morat20 in reply to Patrick Cahalan
                Ignored
                says:

                Don’t think it’s the airlines. NASA has a lot of ass-covering paperwork for that sort of thing. Judging by when it started, and the total lack — to my knowledge — of any travel abuse related scandals — I’d say it’s designed for the every-other-year House Committee Scream Fests wherein our enlightened House Representatives pose for the camera and voters and shake their fists about “waste, fraud, and abuse”.

                Generally aimed at tiny little programs like NASA, NPR, or a host of other small fish.

                You’d think they’d start with the 100+ billion a year programs, but nope. It’s always stuff with tiny discretionary budgets, and generally over tiny little areas of that — like a handful of grants, or whether people are travelling too much or whatnot.

                I suppose, electorally speaking, it’s safer than asking if the Air Force really needs 260 F-22s, or whether the Navy needs that many carriers, or why we’re still paying for an anti-ballistic missile program that doesn’t work, never has work, and never will work.Report

              • Avatar DensityDuck in reply to Patrick Cahalan
                Ignored
                says:

                Morat: Both Standard and THAAD have had multiple successful engagements (and THAAD has intercepted multiple-rv attacks) so I’m gonna have to say you don’t know what you’re talking about. Missile defense is a workable concept and the hardware we’ve got has been proven to work.Report

              • Avatar DensityDuck in reply to Patrick Cahalan
                Ignored
                says:

                Your point would be valid if A: SpaceX were using NASA test facilities, which they aren’t, and B: the NAFCOM study (which is, thus far, THE ONLY THING I HAVE BEEN TALKING ABOUT) had assumed that part of the NASA cost were “amortized cost of multipurpose test facility construction”, which it didn’t.Report

Leave a Reply

Your email address will not be published. Required fields are marked *