Tariq Malik has a piece on the new space prize today, in which he writes:

Former astronaut and U.S. senator John Glenn's 1998 space shuttle seat cost NASA $50 million, and private orbital passengers like Dennis Tito and Mark Shuttleworth have paid about $20 million for jaunts to the International Space Station, McCurdy added. At present, British millionaire Sir Richard Branson's announcement of suborbital flights on his newly christened Virgin Galactic venture will cost around $190,000.

I'd be curious to know where he got the fifty-million number. There is no accepted cost for a Shuttle seat--it all depends on how one wants to do the accounting. I'm guessing that he (or whoever gave him the number) came up with an average cost for a Shuttle flight in the year that he flew (perhaps $350M, itself a contentious number, and probably low), and then divided by the number of crew.

But this is a completely arbitrary way to do it, and in fact extremely overprices it, since it values the cost of delivering a payload bay full of tons of cargo at zero.

The reality is that John Glenn's flight cost virtually nothing, at the margin. They could have flown seat full of John Glenn, or seat empty, and the cost of the flight would have been identical, other than training costs. Unless the services of the Shuttle are "unbundled," there's no definitive way to put a cost on a seat.

Actually, there were costs associated with Glenn's shuttle flight. The training time (facilities and personnel), medical, and administrative costs were all above negligible. Also, the time spent doing Glenn-related things while in orbit must have, of necessity, prevented the crew from engaging in other matters. These costs are probably not as easily determined as the first group. (It's unkind to call it baby-sitting, but it gets the point across.)

Now, these certainly aren't going to total up in the millions. But I'd suspect that the first group would probably be in the several hundreds of thousands of dollars - govt. accounting and all.

It's not just Government accounting, it's all accounting. Deciding which heading to put costs under, how to account for overhead in determining costs, etc... etc... These are not simple tasks, and when they have tax implications it gets even more complex.

If NASA were like a business, then they would have to determine the price of a flight and sell enough seats to cover the cost and turn a profit. So if they filled up the cargo bay with a passenger module and flew 50 people per flight, then I suppose that would be 10 million per passenger based on 500 million per flight, maybe 12 million per passenger to turn a profit. Since NASA is not a business trying to turn a profit,I guess it's moot anyway.

Well and also it's apples and oranges. Suborbital at Mach 3 is not the same as orbital at Mach 25. And also, as Rand points out, the purpose of the SS is not just to get people to orbit. A ride in an aircraft carrier would cost a lot more than a ride on a cruise ship.

I think rather than bashing NASA for being idiots who can't do anything for a reasonable price, it would be better to argue that spaceflight should be privatized because it opens the door to a more diverse and flexible world of access to space. There are many reasons for going up there, ranging from tourism to planetary science research to orbital engineering research and development, military, communication support, and so forth. When the goverment runs the show you tend to get one-size-fits-all solutions to these many different problems. If the field were privatized, there's the hope that there will be many different solutions, each tailored to the needs of a particular market niche, each more reasonably priced for their market niche than the single government solution.

By analogy: some of us are graphic artist Mac people, and some are marketing droid Windoze people, and some of us lucky few are free-spirited technogeek Linux people. It would make no sense and be much more inefficient and expensive to try to have one goverment-specified computing platform for everyone who has computing needs.

In this case, you could reasonably argue that if all Senator Glenn wanted to do was see Earth from orbit, then indeed it would be more efficient use of the taxpayer's dough for the government to buy him a seat on an orbital successor to SS1. On the other hand, if his purpose was to inspect secret military operations in space, then he had to go on a government boat. But the point is: it's always cheaper and more efficient to have customized solutions for significantly different problems, and that's what the private market is good at and the government -- or any monopolist -- is bad at.

So let there be "Mac" spacecraft (SS1?) for those who need it, and "PC" spacecraft (Shuttle?) for those who need it, and "Linux" spacecraft (??!) for those who need that.

Given the size of the shuttle bay and the payload weight limits, what would be a reasonable "max passenger" limit for the shuttle? This should be eminently calculable from public information.

Just by inspection it seems like 'lots', probably 30+. Yes?

I believe $50m probably comes from $500m per flight and 10 astronauts.

http://en.wikipedia.org/wiki/Space_shuttle