Todd Halvorson has an interesting article over at Space.com on the Shuttle and the International Space Station (ISS). It provides a good history and description of many of the issues involved with those programs, and their possible fate.

To summarize, ISS is way over budget, and as usual, NASA doesn't even really know how far over budget it is, or even how much it will eventually cost to get it to the point that it can achieve its scientific objectives. It really is a programmatic disaster of (pardon the phrase) astronomical proportions. (Old NASA joke: "What's the difference between space station and the Titanic? The Titanic had a band.") As a result, the agency is going to have to make some difficult choices in the months and years ahead, and may undergo the most dramatic upheaval since it was chartered in 1958.

But there are some statements in the article that may be misleading, or are not properly given context and explanation.

... in a bid to save money, the agency's $8 billion shuttle fleet will be flying fewer and fewer missions over the next four years, raising questions about the need for four orbiters, three processing hangars and twin launch pads here at Kennedy Space Center.

Well, cutting back flights is in theory a way to save money, but unfortunately, the annual program cost is not proportional to the flight rate, so it's often a false economy. More on that in a minute.

But this paragraph inadvertently points out the real problem with the Shuttle and why it costs so much. It's not because it's a bad design (though there are plenty of things wrong with the design, from an operations-cost standpoint).

It's because we fly it so little, and reducing the flight rate just exacerbates that problem. If we only fly one Shuttle flight a year, we still have to maintain most of the facilities. We could get rid of one of the pads, but then we don't have any facilities redundancy, and the system becomes even more fragile.

We could mothball one or two Orbiters, but that won't save that much either, and there's a minimum amount of manpower that has to be in place to support even one annual launch. You can't call them in to support the launch, and then lay them off for nine months, and hope to have them available the next time you launch. And even if you keep them on salary for the whole year, they forget their jobs if they're not doing them on some kind of regular schedule, a problem that the article does point out, to a degree:

An astonishingly similar set of circumstances was considered as part of an internal NASA study conducted in the mid-1990s. Aimed at gauging the upshot of slashing the annual shuttle flight rate to fewer than five missions, the study came to the following conclusion:

"You don't want the flight rate to be that low because you have a retention problem with the people. You have a skills erosion situation as people leave," Sieck said.

But here's the misleading part:

With nothing other than station missions booked after 2004, the annual shuttle flight rate will remain a constant quartet of missions unless other potential users approach NASA with money in hand.

"It'll be on a pay-as-you-go basis," O'Keefe said. "So the number (of shuttle flights) is going to be based on the demand, if you will, from other science and technology endeavors that we'll be engaged in."

Whether any other "customers" pony up the price of a shuttle launch -- about $400 million -- remains to be seen.

So what they're saying is that four flights a year are planned, and that's all that NASA has a need for, given the reduced activity level on the financially-crippled station. That's one flight per vehicle per year. If the flight rate is going to go up, other customers will have to be found, because NASA doesn't have the budget for it.

Fair enough, but that $400M/flight number is not really valid, for two reasons. First, it is probably extremely low at a flight rate of only four per year (it might be valid for a rate of eight per year). My guess, at that flight rate, the per-flight cost would be more like seven or eight hundred million. That's because it's an average cost. In other words, it is derived by taking all of the Shuttle program costs for a single year, and simply dividing it by the number of flights.

If one is looking for additional customers to boost the flight rate, the proper number to use is not the average cost, but the marginal cost. That is, the actual cost of flying the next Shuttle flight, given that you're already launching some. This is basically the sum of the cost of the propellants and other consumables, the external tank that is thrown away, the cost of refurbishing the solid rocket boosters, and the mission-specific crew training. This is a much lower number than the average cost.

A good guess for this cost would actually be in the neighborhood of a hundred fifty million dollars or so. So if someone wanted to purchase an additional flight from NASA (or some private operator that has taken over Shuttle operations, as is also discussed in the article), they shouldn't have to "pony up four hundred million"--any amount greater than that marginal cost will be worth doing a flight for, particularly since the operator gets the benefit of better utilizing the work force and keeping their skills honed.

Even at the marginal cost, Shuttle is still pretty uncompetitive with any other launcher (with the possible exception of the Titan), so it wouldn't necessarily be unfair competition to allow sales of Orbiter flights. Only if it were to be sold below marginal cost would the taxpayer be subsidizing the system to the detriment of the commercial market. Of course, the main reason that anyone would choose Shuttle over another launcher is its unique ability to carry people comfortably to orbit, and return them.

Thus, this might provide an opportunity for an entrepreneur. Build a thirty-passenger module for the payload bay, buy a flight for a hundred fifty million, and charge ten million apiece. Rough profit, a hundred fifty million, and you're undercutting the Russians by ten million dollars a ticket.

NASA has never been willing to consider such a use for the Shuttle in the past. But in the past, they've never been quite so desperate...

Thank you for pointing out marginal costs. But how can you expect them to understand marginal costs when they can't figure out diminishing returns or efficient use of resources?

Well, all of this is a consequence of the fact that space isn't important or, at least, it hasn't been in the past. All these problems would go away if actually accomplishing things in space became a national priority, but as long as it is basically just high-tech welfare for engineers, there's no point in worrying about things like marginal cost or cost effectiveness. It's effective at creating jobs and as a foreign policy tool, and that's all that matters right now.