In practice this means that the quotes you get back are much harder to assess from a commercial perspective, as you are trying to weed out which companies are underquoting.
Also cost plus contracts almost always end up being cost 'plus plus' if you dig enough under the covers in my experiences in procurement/contract management (there are always hidden fees and profit lines, and too many opportunities for conflicts of interest or to charge more to the open book), so even a low % can just mean more 'hidden' profits. You could write a book on all the ways to extract additional profit from an open book contract.
IMO the best model isn't an either/or approach, it's a mix of models where the right model is used at the right time (including open book, fixed price, rate cards, hybrid contracts e.t.c.) where the contracts are adequately sized & scoped (i.e. several small fixed price contracts with set deliverables which have value on their own rather than one huge contract with one massive deliverable at the end).
The thing with SLS is it was supposed to be low-risk, based on Space Shuttle heritage - it doesn't do much that's new. Despite that, it's on a cost plus contract that's ballooned.
In the process of this, everything is being modernized. This is where all the cost comes from.
HLV would have been a low-risk launcher based on the shuttle. SLSs is a new launcher build from recycled parts.
Building roads and private parking lots have enough in common that I might risk some of my own money on a better way [do something with them] them because once I develop it I can recoup any losses selling to private parking lots. Most things the government does they have a monopoly on though, and thus once a contract is done there is no possibility to recoup any losses.
If you have ever PIed on a government contract you will know that the program manager will almost certainly deviate from what is in the contract. They will call unbudgeted out of town meetings, ask for unbudgeted reporting, demand unbudgeted changes to the deliverables.
As a contractor working with the government, "cost plus" is the only sane option. The program manager is not going to have the bandwidth to renegotiate (and bid out?) in the almost certain event of a change in scope.
The alternative to "cost plus" is defensive billing where the contractor attempts to devine and account for extra non-contracted work. That lead us in the past to $1000 hammers and no-one liked that either.
Insurance companies are in a much better place to price this insurance than NASA, because they have the correct financial incentive to do so well, and no political incentives to do so poorly.
Sure they can, but when they decide to tweak the project the company can tell them "no, that sounds risky".
So what will happen is the contract will end up stating something like all adjustments are priced at cost plus... and NASA is back at square one.
For example with Boeing's recent starliner tests, they would be asking for more money to fly the orbital flight test that they initially promised, because they fucked it up the first time.
Is that really a thing? The risks of adverse selection and moral hazards would seem insurmountable.
Any insurance would be expensive, and would be a cost passed to the government, as any other project related costs are. The insurance provider spreads risks across policies but since all these policies would probably be for government contracts, the costs would all be passed on to government. The government would end up paying for all the policies plus the insurance company profit margin.
The only winner here is the insurance company, maybe the contractors. The government will be slightly worse off.
An insurance provider is a party taking financial risk if something goes bad for someone else. In the cost plus contracting model, the government is taking the financial risk if the project goes poorly for the contractor, that is the government is providing insurance.
> The insurance provider spreads risks across policies but since all these policies would probably be for government contracts, the costs would all be passed on to government.
Of course, there is no free money here. However the government can correctly evaluate the costs of the different proposals under a fixed cost contract model, whereas they are not capable of correctly evaluating the cost of the insurance that they add on top in a cost plus model, because they aren't set up to correctly price the insurance they are selling. This means that in the fixed price model, they can get a much better approximation of the cost in the value/cost equation they are trying to maximize when evaluating bids, and it means that contractors are motivated to provide options that maximize the same, instead of being motivated to provide options that maximize the difference between how the government miss-prices the insurance, and the actual value of the insurance. In the end everyone ends up better off, because there is less waste.
If one doesn't exist, that should just be a sign that the government really shouldn't be providing it either though, because it isn't profitable. The government isn't somehow better placed to insure R&D work (or whatever you want to call the work being discussed) than any other insurance company is.
Cost plus would be "deliver a FTL ship, we'll pay you X * costs, no it doesn't really matter if you go 10x over what you initially quoted us for the work and still don't have the thing". You can see why contractors under that model are willing to make unrealistic quotes and promises about FTL, or for a real world example why SLS is many years and many billions of dollars behind schedule.
For instance, as part of their filing for the HLS competition, SpaceX supposedly had around ~400 pages just discussing cryogenic propellant storage and transfer, along with the associated risks and how they would mitigate them.
It's also a bit of a stretch to say NASA isn't taking on risk in fixed price, as they're still paying large sums of money for each agreed upon milestone. The point is that the risk needs to be shared. Cost+ takes away pretty much all of the risk for the company, taking away their incentive to do their best (especially considering that until SpaceX came around and blew the doors open for smaller companies, there were only a handful of competitors who were all basically the same culture wise).
For an expensive one-off like that there needs to be wiggle room to deal with design uncertainties and the unknown-unknowns.
Fixed price might be more reasonable for smaller less expensive missions where the uncertainties are less and the tolerance for failure is higher.
Cost-plus has been so much more expensive than fixed-cost that NASA could take lots of risks on fixed-cost and still come out ahead.
Now that's something I'm not against.
Plenty of builders can make you a house on a fixed-cost. I used to know one, and every spec house he built was within $1000 of his initial price because he could look at prints and know in an hour how much the lumber, labor, plumbing, electric, cabinets... would cost. He could figure this out even if it was the first time building that print. (spec house was important - if it was custom the owners were use to add $30,000 in upgrades)
SpaceX can quote you a fixed price launch of crew dragon to ISS. However if you want to build a new ship - there are too many unknowns. I'm sure if NASA was content to stick with the opportunity rover design they could have thousands on the surface of mars by now for the budget that has gone into the various programs we have done since then - but we learn a lot more from the new programs that opportunity can't give (perhaps we would know more about mars?).
Most of the things NASA does are things where the risk is far too high for anyone sane to take on all the risk in a fixed price contract. Instead NASA needs to take on the risk in some way. Either that means some form of contract like cost-plus, or a lot of smaller contracts such that everyone can succeed at their contracts while the project itself is a failure (I'll make module X exactly to specs, too bad if you mess up the specs). All forms of such contracts are subject to abuse, NASA needs to figure out how to manage that abuse.