A commenter at this post writes:

When it costs hundreds of millions of dollars to fly a single Shuttle mission, I fail to see the problem with spending another 10 to fix the wiring.

The first problem is a misunderstanding of Shuttle costs. The marginal cost of a flight is not "hundreds of millions of dollars." It's probably somewhere between a hundred and a hundred and fifty million. The average cost is much more, but that's not a useful number, because it can vary so much with flight rate (for example, when the flight rate is zero, as it has been since February, 2003, the average cost per flight is infinite, regardless of how much we spend on the Shuttle program).

The second problem is that, while ten million dollars may not seem like much in the context of a program that costs billions annually, the fact remains that NASA has a finite budget, and ten million spent on one item is ten million less available to be spent on something else, that might be more important. According to the article that the original post linked to, the odds of an uncommanded thruster firing resulting in a catastrophe are somewhere between one in ten thousand and one in a million (it doesn't say if that's on a per-mission basis, or totaled over the next twenty-odd flights). Assuming that those are valid numbers, with any degree of confidence, then the standard way to determine how much we should spend to prevent that event from happening would be to use the expected value of that event (probability times cost). The problem with that, of course, is assessing the value of either the Shuttle fleet, or the ISS, given that current policy recognizes them both as dead ends, in terms of future space policy.

That, in fact, is why I think that the CAIB recommendations should have been revisited after the new policy was announced. If the CAIB had known that the Shuttle was going to be retired at the end of the decade, they may not have recommended some of the more costly (and impractical) fixes for what would then have been recognized as a rapidly depreciating asset.

The subject of Shuttle launch costs has been one long debated. I submit that when talking about a random $10m chunk of cash, that the relevant parameter is "how much money did NASA spend on the Shuttle this year" divided by the number of launches. As you point out, this can be rather large. ($500mill according to one Av week article in the early 90s when the launch rate was decent)

So, you worked at Rockwell Downey?

"the fact remains that NASA has a finite budget, and ten million spent on one item is ten million less available to be spent on something else, that might be more important. "

It irks me that in some arguments, NASA has a finite budget, in others it doesnt. For instance, i have seen people proposing scrapping STS and ISS today, supposedly making many billions available to accelerate VSE and use for other purposes. Then other people jump in and say that the budget doesnt work that way, i.e. if STS and ISS will be scrapped, those budget lines will disappear and NASA will get less money. Which would probably be true.
And now Rand is saying that NASA can only spend extra on some projects if it takes money away from others. So, it follows that NASAs budget can only shrink, except when entirely new projects are added to the budget. Is that why the agency never completes any major projects ( X33, OSP etc ) but continues to propose new ones ?
dont answer :)

anyway, for the most part NASA, its budget and its projects arent particularly interesting or relevant WRT to real space development

I meant a finite budget for the Shuttle. My point is that there are many areas in which one could spend money to improve the reliability of the system, but the budget isn't infinite (particularly on an annual basis), so priorities have to be determined.

And yes, I worked in Downey.

I fear we will have to continue beating the drum to get the shuttle cancelled on schedule in 2010. It is pretty much an academic point at this stage of the debate whether cancelling the shuttle in 2005 would result in more money for other NASA programs. I believe it would, but it is surely moot.