What you are reading is intended for people who have a passing or general interest in either NASA, space, science & technology, or how the government & business interplay in this field. I use a lot of simplifications & generalizations and gloss over some details. What inspired me to write this is the (hopefully) upcoming major launch milestone for SpaceX’s Starship rocket, how I see some of the media coverage around Starship & the SLS rocket launch from 4 months ago, and some conversations I had with people unfamiliar with the space industry. This will NOT be a comparison between SLS and Starship; there are countless articles and YouTube videos that do the comparison and the world doesn’t need any more of those. This is not intended to be unbiased and neutral. I have opinions; I’m sharing them and explaining why I have them. Do your own research and learn about the other side of the argument before forming strong opinions of your own.
With that out of the way, let’s dive into the SLS (Space Launch System) rocket. This rocket first flew on November 16, 2022 and the flight was considered to be a success. It took off from Florida without people on board, put the Orion space capsule into lunar orbit, and Orion took some fantastic future-iconic pictures looking at the Earth and the Moon. This launch was a test mission to see that the rocket system as well as the Orion capsule functioned as intended before allowing humans to ride on the rocket on future missions. This is a NASA rocket, a NASA mission, and the SLS program is a NASA program.
What does it mean to be a NASA program? You might think that there are great NASA factories and workshops where NASA employees build state-of-the-art technologies and then launch them into space. That is wrong. Throw that mental model away. NASA itself and the people who work for NASA don’t actually design or build the marquee stuff you are thinking of.
How things work for the most part is that the engineers and politicians at NASA and in Congress determine that there is some capability or mission that NASA needs to do. They write down the requirements that they have for their needs and then non-governmental companies (you’ll hear the term “contractors” used for these companies) submit proposals based on those requirements. These proposals are usually competitive bids for how they will meet these needs as well as how much money NASA will need to pay them. NASA will then choose a winner based on the best bid across different financial, schedule, and technical considerations. It is the engineers at these contractors who do the actual design and production of the things that NASA wants.
Think of some of NASA’s iconic projects. You are probably thinking of some of the items in the list below. These are things that NASA specified and funded but were built and designed by non-governmental companies.
Project Apollo – Grumman, North American Aviation, Boeing, Douglas, others
Space Shuttle – Rockwell
Hubble Telescope – Lockheed Martin
International Space Station – Boeing (for the American modules)
Webb Telescope – Northrop Grumman
SLS – Boeing
Pathfinder Rover – Jet Propulsion Laboratory (JPL is a federally funded organization but run and managed by Caltech, a private university)
There are engineers who design things and get them built within NASA (with and without outside contractor involvement), but this is primarily not for the multi-billion dollar front-page-of-the-newspaper projects that most people think of. However, this is a simplification. There are examples of manufacturing large, expensive pieces of hardware at NASA facilities. An example of this is propellant tanks for SLS at NASA’s Michoud location. NASA personnel and engineers also do other work related to the projects that the outside contractors build including things like testing, integration, training, monitoring & operations, and project management.
KEY TAKEAWAY – NASA doesn’t do some of the NASA-ish stuff that people think they do. NASA creates paperwork and then companies design and build things based on that paperwork.
So how does NASA know what it wants and needs? How does it know which capabilities it needs to tell the contractors to deliver? Is there a general NASA budget and the people leading NASA are empowered to allocate it in any way they think is the best to further NASA’s goals and vision? No, not really. The people who run NASA do not fully control what NASA must work on.
NASA’s budget is set by Congress, both in terms of the overall size of the budget as well as the specific things that comprise the budget. These are not gentle suggestions from Congress as to what NASA should and shouldn’t do; the budget is passed as a law (a spending authorization) compelling NASA to do the things Congress outlines. To get the budget passed is usually a three-way dance between the NASA Administrator, Congress, and the relevant staffers from the Presidential administration responsible for Space Policy. Some of the items that need funding will be requests from NASA to Congress (“we think that we will get the most bang-for-the-buck in terms of scientific knowledge by sending a probe to Uranus so please add a $1 billion funding line item for that in the budget”) and some of the items will be things specifically that Congress wants NASA to do, regardless of whether NASA wants funding for these items or whether these items are overall beneficial to NASA/country/taxpayers.
The SLS program is an item that Congress put into the budget (specifically the NASA Authorization Act of 2010). In fact, not only did Congress tell NASA that it must obtain the SLS capability, but Congress wrote the spending authorization in a way that also dictates how NASA was to obtain that capability. To put it another way, politicians specified the industrial and technological realities for something that is the domain of industrial and technical experts. There are several problems with politicians dictating technical outcomes; the lack of financial competitiveness in bids and the lack of technical optimization to meet the requirements. Did they at least tell NASA to do something groundbreaking, something that pushed the technological envelope, something inspiring & purposeful? If I’m asking these questions, then you already know that the answer is “no”. Congress dictated that SLS use as much of the existing technology and industry base from the Space Shuttle as possible. To put it another way, Congress told NASA it needed to spend its money on a brand new, very expensive rocket system using as much existing 40 year old (work on the Space Shuttle started in the early 1970s) technology as possible.
Don’t take my word for it, here are some excerpts from the bill (https://www.congress.gov/bill/111th-congress/senate-bill/3729) that created the SLS.
“Requires the extension or modification of existing vehicle development and associated contracts necessary to meet such requirements, including contracts for ground testing of solid rocket motors, if necessary.” … “Ensures the retention, modification, and development of critical skills and capabilities, as appropriate, in areas related to solid and liquid engines, large diameter fuel tanks, rocket propulsion, and other ground test capabilities.”Section 302
“Requires the Administrator, in developing the Space Launch System and the multi-purpose crew vehicle, to utilize existing contracts, investments, workforce, industrial base, and capabilities from the space shuttle and Orion and Ares 1 projects, including space-suit development activities and shuttle-derived and Ares 1 components that use existing U.S. propulsion systems. Specifies the activities that shall or may be discharged by NASA in meeting such requirement.”Section 304
The SLS program was predicted, from the very start, to have the types of problems we’ll see later.
But there must have been a good reason, right? There was some kind of mission or goal that required a massive launch capability that didn’t exist at that point so we need the massive SLS rocket to accomplish that mission? Today, SLS is the planned rocket launch system for the Artemis program of manned and unmanned lunar exploration. Portions of Artemis (so far requiring at least 5 SLS launches) are on NASA’s plan of record and funded, other portions are speculative. Artemis is relatively new (2018) as an official program. Some of the critics of SLS have previously called it the “rocket to nowhere” – a play on words of the “bridge to nowhere” concept, when politicians spend money on infrastructure projects without any justifiable use or value. This is an understandable criticism since for many years after its inception, there wasn’t an official plan of record (the various cancellations, restarts, funding changes, and reformulations of the Constellation program, Orion, and other speculative NASA exploration programs is a story on its own) for anything that would require SLS.
KEY TAKEWAY – Congress made (as a matter of law) NASA fund, in a non-competitive way, a massive rocket program using as much legacy technology as possible.
It doesn’t sound great so far. NASA is obligated to fund a massive (both in terms of cost and performance) rocket program and its hands are tied in the way to design and build the rocket. But look at the bright side, the actual engineering work is being done by companies (not the government) so it will get done economically/efficiently AND we have a history of 40 years with the technology so we shouldn’t have a lot of technical problems, right?
Let’s take a look at how the economics work between NASA and its contractors (this is also applicable for the defense industry, not just a NASA issue). For the most part, contracts have one of two types of structures. One of these types of contract structures is called “fixed price” (I’ll refer to it as FP). In this type of contract, the scope of work is specified and the amount of money that NASA will pay for the work is a fixed number. Contractors must estimate how much the work will cost as a baseline, and then add some dollars for profit, add some more dollars for a safety margin in case things go wrong, all while trying to be competitive with what they think other contractors would bid for the job. For example, if the contractor wins a FP contract for $10 but it actually ends up costing $12 to finish the project, then the contractor will end up losing $2 on the project. If the same job only cost them $7 to complete successfully, they would make $3 more than they thought they would. The entity issuing the contract pays $10 regardless of whether the contractor over-or-underspends. This is simplified because in the real world, there are rules in federal contracting about how much profit percentage a company can actually make. But the point remains; the contractor is on the hook for overages, must eat that risk, and is incentivized to stay on-budget.
The other type of contract structure is called cost-plus (I’ll refer to it as cost+). With this type of contract, the contractor is guaranteed to be paid back whatever they spend (“cost”) and then receive some additional sum of money on top of that (“plus”). For every dollar that the contractor spends, it will get reimbursed that dollar and then get some additional money as well. Let’s see an example of what that looks like where the cost+ contract covers actual costs and then adds 10% on top of the costs. For every $10 that the contractor spends, it will receive $11. For every $100 that the contractor spends, it will receive $110 in revenue. This is a very simplified example and real contracts will have some clauses about what can and can be charged against the contract, performance milestones, etc. But the point remains; the contractor essentially assumes no financial risk for the performance of the project and is going to make a profit on the project regardless of the outcome or their performance on the project.
Now we get to the problem of incentives on cost+ contracts. On a FP contract, the contractor is incentivized to get things right and do them as efficiently as possible in order to get the maximum profit. If you screw up on a FP contract and need to redo some work, it could mean the difference between earning a profit and making a loss. On a cost+ contract, the incentive is quite the opposite. The more money the contractor spends, the more money it makes. The contractor is guaranteed a profit and the larger the sum of its costs, the larger the profit. It can be reasonably concluded that on cost+ contracts, the incentive is actually to work as inefficiently and expensively as possible in order to get the most profit from the contract. If you could do a piece of work for a cost of $1 million and make $100,000 for that work, then why wouldn’t you do that same piece of work for $2 million and make $200,000 profit instead (assuming that is within the contract parameters).
It might not shock you that this is how companies do behave. Companies, even working on noble projects for NASA or important defense research for the Department of Defense, are not operating out of a sense of patriotism or altruism or charity. They operate to make money. Especially the huge multi-billion dollar revenue companies that receive the bulk of NASA and military contracts (Boeing, Lockheed Martin, General Dynamics, Northrop Grumman, Raytheon). These are all publicly listed companies and their #1 priority (not by law, but this is unfortunately the prevailing corporate mindset of our age) is to maximize the price of their stock. In order to do this they need to extract as much revenue as possible from their contracts, they need to hit quarterly corporate financial objectives, and they need to make investors believe that they will keep growing their revenues over time. The executives, high ranking managers, and rank-and-file employees all get some sort of cash bonus or stock grants based on how well their company performs against their corporate financial goals. Milking cost+ contracts at the expense of taxpayers to boost incoming revenue and boost company performance sounds like the way to go to reap maximum financial benefit!
As an aside, why do we have cost+ contracts in the first place? Seems like a very bad deal for the entity who issues and pays for that contract, right? The idea behind cost+ contracts is that cutting edge research and development is difficult and expensive. There a lot of things that are not known when you try to make cutting edge technology. There a lot of things that can go wrong and many potential setbacks along the way. There are a lot of potential financial downsides when you try to tackle a difficult technological challenge. The threat of one exceptionally difficult project, where one thing after another goes wrong, bankrupting your entire company is real. This risk would reduce the incentive for companies to bid on government contracts to work on speculative or groundbreaking projects. Well, the government does want companies to work on challenging technical problems so it needs to provide some incentive to offset the risk. And that incentive is the cost+ contract. Even if things are going wrong, the company will still make money to survive. And if it can survive, it can complete its contractual obligations to deliver the technology it has been paid to create.
Here is a clip of a much more successful, thoughtful, and well-known person than me talking about it.
Let’s ask the question about whether it was appropriate to use a cost+ contract for the SLS. Personally, the answer is a strong NO. The entire SLS program was written into law to use as much existing technology and industry capability as possible to reduce technical risk. This is quite the opposite of the intention of cost+ contracting to reduce the financial risk of developing new things with inherent high technical risk.
How bad has the financial performance of the SLS program been? Boeing is the primary contractor (called a “prime contractor”) responsible for the SLS rocket. It oversees the overall design, development, and production of the SLS rocket. It also contracts out some pieces of work to other companies (called “sub-contractors” or “subs”). Overall, the initial plan for SLS (from 2011) was that it was going to cost $10 billion for the rocket (and an additional $8 billion for the Orion space capsule developed by Lockheed Martin and ground equipment developed by Bechtel and others) and the first launch would be in 2016. To date, the program has cost $24 billion and the first launch was 6 years behind schedule. Even after a successful first launch, there are several pieces of the program that are severely behind schedule and over-budget for subsequent planned launches. The following article is fantastic, but the short version is that the contractor (Bechtel) responsible for the launch tower needs to spend $1 billion (versus a planned $400 million) to redesign the existing launch tower and is years behind schedule – https://arstechnica.com/science/2022/06/nasas-second-mobile-launcher-is-too-heavy-years-late-and-pushing-1-billion/. This is really, really bad overall performance by Boeing. You would think that maybe there would be some sort of penalty or fine for executing on a contract so poorly. Well… no. In fact, NASA rewarded Boeing for its abysmal performance by giving them $260 million in performance bonuses. Instead of not getting any performance bonus, Boeing got paid 90% of the total amount of bonus payment that it was eligible for.
KEY TAKEWAY – The financial incentive for Boeing to design and build SLS is to do it as inefficiently as possible. Boeing is 6 years behind schedule, 30% over budget, and has been given massive performance bonuses on top of that.
We have an incredibly expensive rocket project which was driven by political considerations. It is also years behind schedule and billions of dollars over budget. But it is still an awesome rocket, right?! It kicks butt and can do things that no other rocket can do and does everything better! Right??? Those are very good questions and to answer them, we need to look at what SLS relative to the alternatives and see what kind of capability and value we have versus those alternatives.
On paper, SLS and SpaceX’s Starship are the two best comparison points. But, Starship (at the time of this writing) has not flown even a test mission and might never fly at all for all we know. SLS has flown and successfully put a space capsule in orbit around the moon. It would be intellectually dishonest to compare something that’s already here with something that isn’t ever guaranteed to come. The last thing I will say on the Starship topic is that if it does work as intended, even with half the projected performance at twice the projected price, it will make SLS instantly obsolete across any and all parameters. We need to compare an existing, working rocket system against SLS. Let’s look at Falcon Heavy (I’ll call it FH for short) from SpaceX.
FH first flew in 2018 and has successfully completed 5 missions with 0 failures. One of the most important differences between FH and SLS is that FH is a COTS (Commercial, Off-the-Shelf) product. It is an already existing product and launch service that anyone (private industry, government, domestic, foreign, etc.) can buy right now. It has a known, published price. To use an analogy; FH is like going to a car dealership, pointing at the car you want, paying the price written on the sticker on the windshield, and driving away in your new car. SLS is like wanting a car, enlisting the services of an engineering design firm to design the car from scratch to suit your needs, enlisting the services of a construction company to build a factory to build your car in, and then enlisting the services of a manufacturing company to build your car in the factory.
A FH mission (using the non-reusable configuration to make for a more even SLS comparison, there is also a reusable configuration which is cheaper but can’t launch as much payload) will cost $150 million and be able to launch 140,000 pounds to Low-Earth Orbit (LEO). This comes to roughly $1,100 per pound to LEO, one of the key metrics of the launch industry. In contrast, SLS (using a technical configuration called Block 1 – this is the configuration that has already flown successfully) is designed to put 209,000 pounds in LEO. Both FH and SLS have several other orbits that they quote maximum payload numbers for (and this impacts the $/pound calculation) but I will only discuss the LEO numbers to make the comparison straightforward. To complete the comparison, we need to know what a launch for an SLS rocket costs, and this is not a straightforward answer because SLS is not a COTS product.
There are two ways that we can frame the answer to “what does a SLS launch cost?” The first method is to take the entire total cost of the program and divide that by the total number of rocket launches to get the cost per launch. The second method is a bit different. With this method, the cost per launch is stated as only the incremental cost needed to produce and launch a new rocket. Other costs are not factored in. For example, you need to buy light bulbs to illuminate the rocket factory regardless of whether you build one rocket or 100 rockets. But for each rocket that you build, you would need to buy a new flight computer. So you would not factor in the cost of the light bulbs when calculating the cost per launch but you would need to factor in the cost of the flight computer. It is fine to use either method, but context must be provided as to which method is being used to make it clear what the cost numbers mean.
Unfortunately, the simple answer is that we cannot get a simple answer for what a SLS rocket launch costs, regardless of the method used to get to a cost number. Seriously, go read the Wikipedia entry about the budgeting and expenditures for SLS and the other links below. It is headache inducing. There are estimates ranging from $500 million to $1 billion to $2 billion per launch depending on the methodology and the organization producing the estimate. Let’s be kind to the SLS and use the low end of $500 million per launch. This means $2,400 per pound to LEO on a SLS launch. Again, we are being extremely kind to SLS. When we look at a real-world example, it shows us that over $1 billion per launch could be more accurate. NASA’s Europa Clipper project (a probe to be sent to Europa, a moon of Jupiter) was initially planned to launch on SLS. For various reasons they switched their launch plan to FH and they saved $1 billion on their budgeted launch cost.
We can see that there is a large disparity in value. Comparatively, it costs twice (best case number, not the realistic case number) as much to put a pound into orbit on SLS. But, you may argue, that doesn’t matter since the absolute monstrous performance of SLS provides a capability that can’t be alternatively obtained at any price. The FH heavy rocket can only lift about 66% of the mass that SLS can. That is true, BUT, that reasoning also ignores the opportunity cost of NOT spending money on SLS versus spending the money on SLS.
For the price of one SLS launch, you can launch 3-6 FH missions (depending on which numbers you use). So if you are trying to launch some item into space that weighs 200,000 pounds (more than 1 FH can put into orbit), you would probably be better off redesigning that item to be comprised of 2 components that each weigh 100,000 pounds and launch each component on its own FH mission. You would end up saving about $200 million dollars at a minimum. With that kind of money, you can invest it in adding new functionality (like orbital docking…) to your launch item(s), developing more payloads to launch, or even just not spending the money (heaven forbid, us taxpaying peasants don’t deserve such an option). If the concern is not the weight of the payload you are trying to launch but the distance from the Earth that you want to take it, there’s a solution there too. Launch your payload on one FH. Launch another FH filled only with fuel and no other payload (or to put it another way, additional fuel is the payload). Have your payload dock with the fuel-filled FH in LEO and then that FH can take your payload further out into the Solar System. You still end up saving $200 million dollars versus a single SLS launch to take your payload to the same place. If it is technically possible to break a single, large payload into smaller components or to take a payload far from Earth in several stages, then SLS loses the value of having the capability to launch massive payloads.
There is another opportunity cost that must be discussed in context of the SLS. When Congress made NASA take on SLS as a mandatory project, it was not an incremental addition to NASA’s budget. The budget remained the same and other things needed to be sacrificed in order to fund SLS development. NASA’s budget was $19.3 billion in 2016 and $24 billion in 2022. In 2016, SLS was 10% of the budget. In 2022, SLS was 9% of the budget. What are the awesome things that NASA couldn’t do because SLS sucked up so much money? What kinds of missions to explore different moons and planets of the Solar System didn’t happen? What fundamental research on new propulsion or energy generation technologies didn’t get funded? Which aspects of developing space infrastructure (like propellant depots, habitat modules, etc.) did we fall behind on because of the trade-offs made to fund a rocket to nowhere?
It gets worse. Beyond the opportunity cost of not doing other awesome stuff, the terrible spending decision on SLS is now driving further terrible government spending on expensive space projects without a good purpose. Specifically, the Lunar Gateway. The Lunar Gateway is envisioned as a space station that would orbit the Moon. It is being discussed as something that would aid us in going back and landing on the Moon and can only be successfully deployed by SLS. How convenient. Instead of cancelling SLS, the government decided to waste billions of dollars more on a project that would retroactively provide additional justification for the existence of SLS. Currently, the total estimated cost for the Gateway is unknown but nearly a third of a billion dollars has been allocated for studies and proposals about it (it must be noted that this project is not on the plan of record or funded, it is still speculative and there is a possibility it does not get pursued). But there’s more. Some folks looked at the proposal and determined that the Lunar Gateway concept would actually be a detriment to lunar exploration. It would cost more money, take more energy, and take more time to send something from Earth to the Gateway, then down to the Moon, then back to the Gateway than just straight from the Earth to the Moon and back. This is how much your government respects your tax dollars. Thanks, serf, now back to work!
KEY TAKEWAY – Launches on SLS are incredibly economically inefficient and SLS has wrecked and continues to wreck NASA’s budget to do other, useful projects.
Why should you care? Two reasons. The first reason is that the government is massively wasting billions of your tax dollars on something that is a net detriment to space exploration while filling the coffers of Boeing and the other contractors working on the SLS program. This should make you angry as a taxpayer. You should be angry at the politicians (this is not ideological, both parties are the problem), the funding & contracting mechanisms of our government, and the lack of integrity by the people involved. The second reason to care is that the media is not doing an adequate job in educating the people about the situation. There are certainly space/industry journalists and outlets that cover SLS from every perspective (just look at all the links I’ve provided for proof). But you should be angry that the nightly news on your local news station didn’t tell you the full story behind those wonderful Orion images that were taken from lunar orbit. What kind of society do we live in where something like SLS, which is a prime example of political, governmental, industrial, and technological failure, is sold to us as a success by politicians, bureaucrats, and the media?