3 ms·
I disagree. Cost-Plus might not be the right model, but fixed price only works if all the risks are known and so easy to account for. Plenty of builders can m
by bluGill 5y ago
I disagree. Cost-Plus might not be the right model, but fixed price only works if all the risks are known and so easy to account for.
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.
- gpm 5y agoNASA doesn't need to be in the business of managing that risk and the abuse that comes with it. NASA can pay a fixed cost equal to the expected cost of the project, and the contractor can buy insurance if they want insurance against cost overruns. 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.
- teruakohatu 5y ago> NASA can pay a fixed cost equal to the expected cost of the project 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.
- gpm 5y agoYou don't get to change the deliverables of cost plus contracts without re-negotiation either. That's not the risk that is being avoided here, the risk is that the company is going to decide to ask for more money for the same service they initially promised. 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.
- NovemberWhiskey 5y ago>Insurance companies are in a much better place to price this insurance than NASA Is that really a thing? The risks of adverse selection and moral hazards would seem insurmountable.
- bluGill 5y agoYou can get insurance on anything you want. However the more unknowns there are, the more they will charge. You can get fire insurance that will cover even if you burn your house down - but it will cost you (not having insurance and burning your house down might cost $.01 more than buying insurance)
- gpm 5y agoYou can negotiate insurance for just about anything (there are sometimes laws against things like taking out life insurance policies against a third party, but none that would apply here). No doubt the insurance company would insist on various forms of audits, similar to what the government insists of for cost plus contracts today. Probably the most successful contractors wouldn't pay an insurance company for this, because it's a stupid model for contracts that wastes resources, but it wastes no more resources than having the government play the role of the insurance company.
- teruakohatu 5y agoI think you need to articulate why you think the govt. is forced into this insurer role. They are no different to any large private buyer. 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.
- gpm 5y ago> I think you need to articulate why you think the govt. is forced into this insurer role. They are no different to any large private buyer. 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.
- trothamel 5y agoSpaceX did bid fixed-price for HLS, the moon lander they're working on. That project will required developing refueling from an on-orbit propellant depot to succeed. 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.
- nickff 5y agoNobody in their right minds would have touched SLS with a fixed-price contract; NASA is too involved in the design, construction, and testing.
- sidewndr46 5y agoNot just heritage, the SLS is basically what you get if you move the Space Shuttle engines onto the propellant tank and remove the orbiter vehicle. Then discard the reusability requirement and you've got SLS. In the process of this, everything is being modernized. This is where all the cost comes from.
- bluGill 5y agoSpaceX only did that is they wanted to do it anyway, so they were willing for the risk
- throwawaylinux 5y agoGreat. Seems better to shift the risk evaluation to the company which is full of experts of its own capabilities.
- bluGill 5y agoOnly if you can do that. Most of the large projects the government does nobody else wants. There is no market for aircraft carriers other than the US government (even if there were not export laws), likewise most space exploration, nobody other than NASA is going to buy your mars rover. 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.
- uberman 5y agoThis 100% and it is not just super risky things like "exploring space". 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.
- Aperocky 5y ago> stick with the opportunity rover design they could have thousands on the surface of mars Now that's something I'm not against.
- bluGill 5y agoWhat we could learn from that is a lot less than what we learned from all the other things we have done instead. Eventually you have seen that type of rock before and so all you know is there is a lot of it - which we already have reason to suspect.
- dotnet00 5y agoBut isn't accounting for risks and having contingencies to minimize their impact part of good engineering practice? Eg. when designing a new rocket, you know that the engines might have issues causing delays, you can't predict exactly what issues but you can estimate how much extra money and time you could lose there. 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).
- Closi 5y agoThe counter-problem is that a cost-plus contract works the opposite way - you are incentivised to account for none of the risks in your initial bid to make your quote competitive, and the sales strategy is to 'land and expand' (i.e. get the initial contract and then expand it's value - sales teams will openly talk about land and expand behind closed doors). 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).
- lamontcg 5y agoStrict fixed-pricing seems like a recipe for something like JWST to fail miserably in the future. 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.
- cryptonector 5y ago> I disagree. Cost-Plus might not be the right model, but fixed price only works if all the risks are known and so easy to account for. 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.