WSJ Events Sep 20, 2026 23m 11m saved
With Jared Isaacman, Administrator of NASA
Three quarters of the people doing NASA's work were not NASA employees. They were rented from staffing agencies, including the flight controllers the astronauts talk to in Houston.
The usual reading of NASA's slow pace is that safety reviews and congressional politics hold it back. Jared Isaacman says the binding constraint was the org chart: a civil-servant headcount cap imposed about 30 years ago pushed the agency into renting most of its technical workforce by the hour, and he is hiring those people back onto the payroll.
"It was 75% of the workforce we were renting indefinitely because 30 years ago somebody instilled these arbitrary civil servant caps on our various centers, as opposed to anything statutory."
Jared Isaacman, Administrator of NASA, on WSJ Events, built the payments company Shift4 starting at 16, co-founded an aviation training company, and has flown to low Earth orbit twice on SpaceX vehicles, which makes him the rare agency head who has been the customer and the passenger before becoming the buyer.
The full interview is covered here so you can skip it. 23 minutes of audio, 12 minutes of reading.
Here are the 8 calls that matter.
Key Takeaways
75% of NASA's technical workforce was rented from staffing agencies, and he is bringing it back in house
Mission control had one civil servant in it; everyone else was contract labor
The lunar Gateway is dead and a base on the lunar surface replaces it, because a station in orbit would have watched China work
NASA used to launch every three and a half months across its own designed programs; one launch a year is now treated as aggressive
The requirements themselves were the schedule problem, not the contractors — a $2B rover as a service became an order for 10 rovers
The only genuinely new technology in the moon architecture, he says, is transferring cryogenic fuel between two Starships in orbit
The lunar South Pole is the size of Washington DC on a moon the size of Africa, and the landing sites are finite
A lunar lander costs about $200M, not $10B, which is what he says makes the program survivable across administrations
Risk tolerance had moved too far toward zero for a program that puts people on rockets at 25,000 miles per hour
1. In The Weeds On Purpose
Isaacman was asked why he makes so many calls himself rather than delegating them, given NASA's size. He answered with an argument about information rather than control: a founder in the early days sees everything, and a scaling organization sees progressively less.
The decisions get worse because the information gets thinner as the organization grows
I think it's imperative to have a pulse on as many aspects of the organization as possible.
Jared Isaacman
He said he is not in the job for the title, and named the deadline he thinks he is working against.
He says a second space race is running, and that it will be decided in months
The Chinese will absolutely put their astronauts on the surface of the moon, do what the Soviets never could, and I think the implications of that will be felt around the world.
Jared Isaacman
His measure of how far the agency's focus has spread is the length of its own authorizing law. The 1965 act ran about five pages and told NASA to beat Russia to the moon.
The authorizing statute has gone from about five pages to 155
It's 155 pages now. We're doing a lot of things in a lot of direction.
Jared Isaacman
Asked for the one thing he has stripped out, he went to the pace of the 1960s: between the uncrewed Artemis 1 flight and Artemis 2, which carried four astronauts around the moon on April 1, 2026, NASA of the Gemini era flew 12 missions.
2. Gateway Cut, Base In
The clearest program cancellation is the lunar Gateway, the space station NASA planned to put in orbit around the moon.
The Gateway was billions of dollars spent on watching from above
For example, the lunar gateway — we were going to build the space station above the moon so we could look down and watch the Chinese having all the fun on the surface — to actually building a moon base, which is what we were intended to do.
Jared Isaacman
It was late and expensive, and it existed to assemble partners
That was a program that was costing billions of dollars, way way behind schedule.
Jared Isaacman
His account of why it survived as long as it did is that it produced announcements, what he called flags for the press release, rather than astronauts on the surface. He said the orbit it used carried a fuel penalty, added a step and was less safe, and that it was designed around bringing international partners together rather than around the mission.
3. 75% Of NASA Was Rented
The moderator said reporting on Artemis 3 and 4, scheduled for 2027 and 2028, turns up widespread skepticism about the dates, and asked what he knows that the skeptics do not. Isaacman's answer moved to the workforce.
Most of the people building NASA's rockets were on somebody else's payroll
It was 75% of the workforce we were renting indefinitely because 30 years ago somebody instilled these arbitrary civil servant caps on our various centers, as opposed to anything statutory.
Jared Isaacman
The example he gave is the one room the public knows by name.
Mission control was contract labor
So when the astronauts called home to Houston, it was from staffing agencies.
Jared Isaacman
The fix is hiring them in, and paying for outcomes rather than hours
So we're bringing that all back in house.
Jared Isaacman
He allows that renting capacity has a real use — surge work, or a technical problem the agency meets once. What he objects to is renting indefinitely. He also put a number on how far the launch cadence has fallen.
The historical rate was a launch every three and a half months
Throughout all of NASA's history on NASA design programs, we averaged a launch every three and a half months.
Jared Isaacman
Against that, he said, a plan to launch once a year is treated inside the industry as impossible.
4. 10 Rovers, Not One
Pressed on the hardware, which includes two untested landers from Blue Origin and SpaceX and new suits from Axiom Space, Isaacman said the delays came from what NASA asked for, not only from who was building it. The replacement space suits had grown requirements he called impossible, including EVAs far longer than anything flown, and were stripped back.
Requirements jumped to the finished state instead of building up to it
Let's get back to doing things in a logical iterative way.
Jared Isaacman
The rover contract is his worked example of the change.
One $2B rover as a service became an order for 10 rovers
We were going to go with this $2 billion rover as a service, jump right to the dream state. Now, we're buying 10 rovers, progressively getting better.
Jared Isaacman
On the landers themselves, he said the contracted vehicles will put enough tonnage on the surface to take the base from an outpost to a small city quickly, and conceded that something closer in size to the Apollo lunar module would have been the sensible intermediate step.
5. Refueling In Orbit
Asked how a schedule gets met without pushing the risk too far, Isaacman said the agency's risk posture needs to come back toward the middle, because it had moved to something close to zero tolerance.
Human spaceflight carries risk that cannot be engineered to zero
If you're going to go to the moon, come back safely, and have people riding rockets at 25,000 miles per hour, there is some degree of risk.
Jared Isaacman
On the technical side, he narrowed the novelty in the whole moon architecture to one item: transferring cryogenic propellant from one Starship to another in orbit. Everything else, including Blue Origin's lander, uses capabilities already flown.
One unproven step, and the company doing it has every reason to solve it
So I have no doubt, I'm highly confident that they will solve the problems of rapid reusability and cryoprop transfer, because again it's critical to the entire strategy.
Jared Isaacman
He credited Elon Musk with the analogy he uses for why reusability matters, and described the fuel transfer in aviation terms.
The comparison is throwing away the airliner after one trip
How much would it cost to take your family to Disney World if you threw the 737 away?
Jared Isaacman
In operational terms it is aerial refueling, moved to orbit
They're basically doing the equivalency of air-to-air refueling.
Jared Isaacman
Asked when the test flies, he would not fix a date, and said SpaceX learns from each launch rather than working to a calendar.
The propellant transfer test is late this year or early 2027
Oh, they'll do prop transfer tests in like very early 2027, maybe end of this year.
Jared Isaacman
6. What NASA Must Build
On where to draw the line between buying a service and building a capability, Isaacman used technology readiness as the test. Where NASA is one customer among many and the product is close to a commodity, it should buy at a fixed firm price and let competition drive the price down.
Launch, observation and communications are real markets, so NASA buys them
So launch, observation, communications — hands down, those are real markets.
Jared Isaacman
The other end of the scale is where he thinks the agency justifies itself, and his example is nuclear propulsion, which he says has no commercial buyer because the same industry can sell power on Earth at scale.
Nobody builds three one-off nuclear spaceships when terrestrial power demand is this high
Why would you deviate from that to build the three one-off nuclear spaceships for NASA to go explore the outer solar system moons when you could be printing these things off and printing money for a long time.
Jared Isaacman
The agency's job is the part nobody else can do
That is what your space agency is at its best, doing what no other agency, company, organization or nation is capable of doing.
Jared Isaacman
Asked whether NASA has become too dependent on its contractors, he said the mistake was not commercial partnership itself. Boeing, McDonnell Douglas and Northrop Grumman were there in the 1960s and are still in the business. What went wrong was spreading the money across everyone and then labeling the result commercial.
Handing an impossible problem to industry and calling it commercial is unfair
And that's unfair, because now you're shifting that near impossible burden from a government agency with a lot of resources and the lowest cost of capital onto industry with the idea that private capital will solve this.
Jared Isaacman
He said the moon base is a case where the commercial model does work: NASA put out a demand signal to build landers and rovers, several of the companies answering it are publicly traded, and the demand signal is what underwrites their investment.
7. The Pole's Parking Spots
On why a permanent surface base rather than visits, Isaacman gave two reasons. The first is that the moon is where the technology for Mars gets proven.
The moon is the proving ground three days from home, before a nine-month trip
Because it is the technical proving ground 3 days away before you go to Mars.
Jared Isaacman
The second is geography, and it is the most concrete thing he said all session.
A moon the size of Africa, with a usable region the size of Washington DC
So for everyone's benefit, the moon is the surface area of Africa. The South Pole, which is where the water ice is and the crater ridges that give you near eternal access to solar light, is the size of Washington DC.
Jared Isaacman
The landing sites big landers can use are countable
There's only so many roads and parking spots that are suitable for big landers to come down.
Jared Isaacman
China's plan was a surface base with a Russian-built nuclear plant
China had the plan to go there, build the base, put a nuclear power plant there with the Russians, and there were certainly folks inside NASA that understood the importance of it, but the direction was to do the exact opposite and just build a base above them in orbit and watch them do it, which made no sense whatsoever.
Jared Isaacman
8. Cheap Enough To Survive
Asked whether the base survives a change of administration, Isaacman answered in one word: 100%. His reasoning is partly political and mostly about unit cost.
The strategy has support on both sides, and he says it cannot be unwound
There is no undoing these decisions.
Jared Isaacman
A lander is a $200M decision, not a $10B one
They cost $200 million. They don't cost $10 billion each.
Jared Isaacman
That is what makes it adjustable rather than cancelable
It's very easy to dial it up or down.
Jared Isaacman
The same logic runs through the science program, he said: an administration that wants to spend less takes a mission out rather than killing the architecture. He named Enceladus as the kind of thing that would go. His contrast is with what he called the too big to fail, too costly to succeed approach, which he says left previous administrators with nothing.
Bonus Insights
The audience asked the question about culture, and he rejected the premise
Someone in the room asked from the floor how a businessman changes the culture of a government agency to move at startup speed — a problem, he said, that everyone in the audience is wrestling with. Isaacman said that is not what he thinks he is doing.
But just to be clear, right, I'm actually just trying to restore NASA back to its glory days.
Jared Isaacman
It's actually just to dust off the playbook that worked so well from decades past and course correct all of the misdirection, the external impositions that have taken NASA off its course
Jared Isaacman
He says the workforce did not resist
Isaacman interrupted the moderator's next question to add that there has been no cultural fight, because the reforms point at the work people joined for.
the workforce has responded so incredibly well because that's why they came to NASA, to change the world in air and space
Jared Isaacman
On continuity, the answer is speed rather than succession planning
Asked how he makes any of it outlast his own tenure, he gave a cultural rule and then went back to the programmatics.
don't let any problem slip to tomorrow that we can solve today
Jared Isaacman
The moderator explained the fuel transfer for the room
Partway through, the moderator stopped to tell the audience what cryogenic propellant transfer is: a Starship-to-Starship fuel transfer SpaceX is developing so Starship can fly deeper space missions, including a NASA moon landing in 2028.
Isaacman's bottom line is that NASA's schedule problem was never the contractors: it was a rented workforce, requirements written for a finished state instead of a first version, and a program in lunar orbit that kept American astronauts off the surface — and he is betting that fixing all three gets US boots to the South Pole before China takes the parking spots.
Products, Companies & Tools Mentioned
NASA (The agency he runs. He says its authorizing act has grown from about five pages in 1965 to 155, and that 75% of its workforce was rented from staffing agencies until this year)
SpaceX and Starship (The lander for Artemis 3, and the source of the one unproven step in the architecture: transferring cryogenic propellant between two vehicles in orbit)
Blue Origin (Building the second lander; he says it leverages capabilities already proven in space)
Axiom Space (Building the new space suits in Houston, after NASA stripped back requirements he called impossible)
Artemis and the lunar Gateway (The flagship program, and the orbiting station he canceled in favor of a surface base)
Shift4 (The payments company he founded at 16, and the experience he draws on for the argument about information and organizational size)
Boeing, McDonnell Douglas and Northrop Grumman (His evidence that NASA never worked alone: the 1960s contractors, all still in the business)
Project Gemini (His pace benchmark — 12 missions flown in the 1960s in the time between Artemis 1 and Artemis 2)
Books & Resources Mentioned
The national space policy (The document he says sets the priorities he has turned into NASA's ignition priorities, and the source of the two-and-a-half-year deadline he is working to)
The NASA Authorization Act (His measure of lost focus: about five pages in 1965, 155 pages now)
Listen to the full episode
🔴 YouTube
Watch the full episode:
More Jared Isaacman interviews and podcast appearances
If this was worth your time, send it to someone closer to the industry than you are.
Get the latest market chatter as it happens: