November 22, 2004
Alex Tabarrok Argues Commercial Space Flight A Distant Prospect
Alex Tabarrok of Marginal Revolution argues in a Tech Central Station article that commercial spacecraft safe enough to enable a large commercial space tourism industry are still a very distant prospect.
Since 1980 the United States
has launched some 440 orbital launch rockets (not including the Space
Shuttle). Nearly five percent of those rockets have experienced total
failure, either blowing up or wandering so far from course as to be
useless. The space shuttle has a slightly better record of safety -- it
was destroyed in two of 113 flights. There are lots of millionaires
willing to spend one or two million dollars for a flight into space but
how many will risk a two to five percent chance of death?
It is true that we have been "learning by doing" or in this case by
learning by exploding. In the 1960s the risk of failure was a stunning
12%. As in other industries, learning by doing reduced the failure rate
dramatically over the first units but more slowly thereafter. In the
1970s the failure rate dropped to 5.2% but nearly thirty years later
the failure rate for rockets still hovers between four and five
percent. We can expect similar slow and steady improvements in the
future but there is little reason to expect dramatic improvements in
If progress continues at the same rate as it has over the past 30 years
how long will it take to achieve a level of safety of say a 1 in 10,000
chance of failure? Note that in comparison to other means of travel
this is very, very dangerous. Commercial airlines, for example, have a
fatality rate of .2 fatalities per 1,000,000 departures or a 1 in 5
Projecting from the historical rate of progress in improving rocket launch safety Alex shows that if that rate of improvement continues then even in the 22nd century space launch will not become as safe as air flight is today.
Some space enthusiasts argue that space flight safety improvements are retracing the same pattern that aviation safety improvements followed,. However, I've done my comparison of aircraft and spacecraft safety records. By my estimation airplanes in 1938 were orders of magnitude safer than spacecraft today.
But let's go back even further to look at aircraft safety in 1938.
That's when some US government agency was created that started tracking
aircraft safety. It is not clear from the table what kind of fatal
accident rate measure they were using. But compare the 1938 rate of
11.9 to the 1950 rate of 5.0 and the year 2000 rate of 1.1. The 1938
rate of fatal accidents was about an order of magnitude higher than it
is now. But its still more than two orders of magnitude lower than the
fatal accident rate of the Space Shuttle.
1938 was 35 years after the first aircraft flight of Orville and
Wilbur Wright on December 17, 1903 at Kitty Hawk North Carolina. Manned
space travel began on April 12, 1961 when a Soviet air force pilot,
Major Yuri A. Gagarin, made an orbit of the Earth. So manned space
travel is over 40 years old. Space travel into Earth's orbit is orders
of magnitude more dangerous after 40 years than aircraft travel was
when it was only 35 years old.
Note that comparisons between airplanes and spacecraft using the data I cited are difficult becuase a lot of airplace safety statistics are in accident rates per million miles flown. This is not useful for spacecraft that get up into orbit and then go around the Earth many times without and ending up where they started each time. Also, Alex's use of fatalities per departures is a more useful measure. However, even there it is not clear to me whether the 1,000,000 departures are departures of humans or departures of aircraft. So I'm not sure if he is doing an apples to apples comparison with rocket and space shuttle launches.
Cronaca blogger David Nishimura points out that the death rate for climbing Mount Everest is comparable to the death rate from rocket launches. Rand Simberg responds that SpaceShipOne supposedly changed the rate of rocket launch technology advance. Alex is having none of this argument. Alex says that refinements in existing approaches using existing state of the art in materials and design just can not begin to close the gap in space flight safety as compared to aircraft safety.
I admire Rutan and I have little doubt that he has made significant advances in rocket design but what I showed in my article was that safety could have improved by a factor of ten or even 100 and rockets would still be too unsafe to support a large tourism industry.
What's so great about space tourism anyway? Even though an increase in rocket safety of a factor of ten is not much when considering the safety of large numbers of people it is very significant when thinking about satellite launches or temporary low-orbit launches. A reduction of risk of this amount means much lower insurance costs that will open up space to new private development.
I am with Alex on this point. Newer rockets have been designed in recent years and have unexpectedly blown up on launch. Rutan's accomplishment is not as radical as some media reprots present it for a number of reasons. First of all, whether he has designed a safer spaceship is will not be proven unless and until it has flown hundreds and even thousands of times without mishap. Also, and very importantly, SpaceShipOne does not do that much. It can not achieve orbital velocity or decelerate from orbital velocity. In my view the Scaled Composites SpaceShipOne flight was important because it demonstrated the potential for prizes to spur innovation. It also opens up the possibility that dangerous orbital spacecraft can be designed and built for much lower costs than NASA and big aerospace companies typically spend.
So then will there be little space tourism in the 21st century? Will you never be a space tourist? Nanotechnological advances may provide new materials that will allow very safe spacecraft or an even safer space elevator to be built. However, even if the answer to the first question remains "No" for most or all of the 21st century and even if you decide you do not want to risk travelling to space in spacecraft that are less safe than today's jumbo jets that does not mean you will never travel into space. If you want to live long enough to travel into space your best bet is to support the development of Strategies for Engineered Negligible Senescence. Donations to the Methuselah Mouse Prize will increase your odds of living long enough to still be around when space flight becomes cheap and very safe. If you want to be able to do more in your life support rejuvenation for life extension. My guess is that life extension will become possible long before space flight becomes as safe as commercial aviation is today.
Let's also not forget that during Rutans second (third maybe?) flight to the all-important 62 mile range, the pilot experienced some serious unexpected vibration for a few minutes, and fortunately was able to back the flight in to free fall and land without incident. There were considerable amounts of discussion if they would be able to go back and try it again. Either way, the flights were by no means flawless, and any extrapolated "odds of surviving" statistics would be projections at best, not hard fact.
This discussion of safety misses utterly the mark.
Current .gov tech made spacecraft/launchers are abysmal white elephant constructs where failure IS an option. Of course they are unsafe. One barely has to look at the shuttle record to realize they suck -- and do so because they originate out of a non-competitive .gov environment where statistical quality assurance and Deming management methods are all but non-existent.
If we are going to have a space tourism industry, you can bet your short hairs that safety will become paramount for the same reason airplanes have been made safe -- namely you go out of business if you detonate on the way to orbit.
By your logic the C-17, the C-5, the F-16, and countless other government-owned planes ought to be crashing every day. Yet they rarely crash.
I've worked in the aerospace industry at a site that built stuff for commercial aircraft, military aircraft, rockets, and space probes. The same engineers moved between the government-funded and private industry projects many times in the course of their work lives. I didn't see huge differences in quality of work for the various projects.
Commercial launch companies selling launch services to satellite operators do not go out of business if they lose a satellite during launch. They just pay large insurance premiums based on their past rates of failures. So why don't the commercial satellite launchers have a lower rate of failure?
Market ideology is not a substitute for huge amounts of technological advances.
Didn't expect the post, but yes, they do crash. All the time, and at quite high rates, and reported on the evening news with regularity. Miramar Airbase here in good ol San Diego, must lose anywhere from 5-10 aircraft a year. Given their takeoff rate, it doesn't take a PhD in math to realize that commercial carriers kick the ass of the military in pure takeoff and landing safety.
There is no such thing as a commercial satellite launcher, there are .gov contractors that are supplementing their income by selling into the slightly deregulated space market. The .gov externality is enormous. Because of said externality, there is no improvement in quality.
And yes, the quality of product between .gov efforts and private efforts is enormous. Compare 777 creation and market deployment with any of the absurd cost-overrun, problem-wrought fighter aircraft built by the .gov side of same said manufacturer. If that's not a difference in quality I don't know what is. Of course, one can also just look at how things were done in the Skunkworks at Lockheed to realize that the normal processes suck bigtime.
What the author of your referred article fails to understand is that safety is a design engineering goal that by quality standards must be designed into your product prior to building it. It is not a function of technology but rather a function of design planning.
Prior to Deming and statistical process management, which showed how to really do things right, I would have agreed that technical advance was necessary for safety. That is no longer the case.
So, no it's not market ideology -- in order to make money you have to be safe, especially if there is no .gov subsidy hanging out in the wings.
All my best and thanks for the great site!
I haven't seen any reports yet of C-17A aircraft crashing. Maybe one has gone down. But I've yet to hear about it. Just did some googling and couldn't find anything on C-17A crashes. There are thousands of C-17A flights every year. The plane is performing very well.
The C-17A had big cost overruns. It had lots of design requirements changes. But it also had a very demanding set of requirements in the first place. It is the only aircraft flying that I"m aware of that has flaps (stronger than the average flap) that go down behind the regular jet engine to give the aircraft more lift on take-off.
Even the B-1B has a better safety record than rockets. They've lost a few B-1Bs. One was due to a flock of birds flying into engine inlets.
All the bombers have much better safety records than rockets. Why isn't the B-2 crashing left and right? It has a radical design. It was funded by the government. Yet it is hihly reliable.
Fighters crash because they are hot rod aircraft. The design trades off reliability to get more performance. The pilots fly very aggressively. Yes, some are going to crash. But fighter jets are orders of magnitude more safe than rockets. F-16 fighters crash at varying rates from year to yet but are crashing at well less than 10 per 100,000 flight hours. They are an aging aircraft. They have a single engine and so have no spare when an engine fails. Half their crashes are due to pilot error. But the margin of error a pilot has on a fighter jet doing aggressive practice flights is enormously less than the margin of error a pilot has on a jumbo jet. The job of flying an F-16 or other fighter with high g's and high supersonic speeds is just a whole 'nuther ballpark than what commercial pilots experience.
Yes, commercial development projects are more efficient. But rockets are made by many companies and all rockets are orders of magnitude less reliable than military aircraft made by those same companies.
Why does everyone compare the rates of improvement from when the wright brothers first flew? Isn't it just as realistic to compare the rates from when the Mongolfier brothers first flew (1783)?
If you compare the rate of advance from the FIRST flights, the current spaceship improvement rate doesn't look so bad.
The early balloons are a MUCH better analogy to early space launches. They were single use, throw away craft. They even sent up animals first as a test.
Note: Flight didn't really start getting much better until the wright brothers, ie. a completely new approach was tried.
Note 2. Exactly the same point is made about that old comparison between cars and computers, where if cars had improved as much as computers had since 1950 a rolls royce would cost $5 and drive around the world on one litre of fuel.
This ignores that motorized land transport was quite mature in 1950. If you compare the improvement in computers to that of the early steam trains, then the modern car IS super cheap, super fast and super thrifty.
You have to get the analogy correct, if one side is wildly better than the other, perhaps you've given it an unfair advantage.
We are failing to communicate. The issue is not that rockets are orders of magnitude less reliable than aircraft.
Why are they that way? Because safety and reliability WERE NOT DESIGNED INTO THEM.
We strongly disagree on the meaning of reliability. Reliability to me is six-sigma plus. Not a single damn aircraft of the military is anywhere near that limit. Commercial aviation routinely runs between 6 and 7 sigma e.g between 1 in 1 million and 1 in 10 million give or take.
The analogy I was trying to get across is that compared to the commercial sector, the military sector is fundamentally unsafe. Commercial aviation would not work with the safety record of military aviation. Period. Why would ANYONE expect space flight, if done with the track record of .gov, to be safe? It can't be done that way. It has to be done by choice, by incentive, by plan -- ie. by those with the vested interested in safety so it's a DESIGN CHOICE at the beginning.
That is the difference between where we are and where commercial folk are going to take space tourism. You don't need technological breakthroughs, you just need solid contextual engineering skills. Something that is not usually taught but is nonetheless the key to success.
I hope I made it across the divide to you ... if not, to you the last word, and we shall agree to disagree.
Again, thanks for the forum.
If you use the 1783 balloons as a starting point for measuring air flight safety then we have much longer to wait for spacecraft to become safe. However, lighter-than-air balloons are as different from winged aircraft in terms of principle of operation as winged aircraft are from rockets.
However, how can we even compare balloons? Is much known about their safety record? I wasn't even able to find much about fixed wing aircraft accident rates before 1938.
I understand what you are saying. I just disagree with you on engineering grounds.
DESIGN CHOICE, as you upper case it: Sometimes you can make some choice to set some design goal and it is just impossible to meet. You can tell me that the commercial industry can achieve some x sigma record of reliability if it just sets out to have that as a goal. But, no, that is not always the case. Sometimes the materials available, the modelling tools available, and the fabrication technology available just makes the goal impossible to reach.
I write software for a living. I've worked with digital circuit designers, physicists, analog designers, and specialists in control laws, chemistry, and other disciplines working on really complex hardware and software designs. I have code executing in orbit around Saturn and in a lot of other places. I know all about the limits of the human mind's ability to design without error and have seen all sorts of unexpected design flaws pop up in complex systems. There is a substantial error rate in software writing. Ditto for digital circuit design. Ditto for some other types of designs. You can't always just declare some design goal for reliability and achieve it.
Some types of gadgets are simple enough and the demands to be placed on them are sufficiently unchallenging given the state of the art that they can be designed and their manufacturing processes can be run to achieve incredibly high levels of predictability and reliability. But that is not always the case. There have been plenty of examples for decades where in some commercial industry some process or design wasn't as reliable as desired and really smart people couldn't figure out how to make it more reliable. Customers would have paid more for higher MTBF. But no one could figure out how to do it. That is the case with rocket engines for example. The understanding of the fluid dynamics and other stuff just isn't good enough. In other cases no one can figure out the physics well enough to design something well enough to work even with low MTBF. That is the case with scramjets.
Yes, I have read Deming's Out Of The Crisis and books about W. Edwards Deming. I understand your point. I just disagree with you on engineering grounds. I do not think you realize just how poorly understood some physical processes are and how hard it is to come to a sufficient understanding of those processes to achieve the level of control of those processes that will yield the high level of reliabilty that we see with other classes of devices.
My co-blogger David Masten has written about this. He says that the data is based on disposable vehicles. There are always tradeoffs between cost and safety, and disposable vehicles naturally tend towards the cheap end, since you are only using them once. Permanent vehicles for space travel will choose a different point on the spectrum.
Of course, if people have indefinite lifespans they may demand much higher levels of safety than even those offered by contemporary commercial aircraft.
Interesting. Such a static view. That's not how it really works. Failure to achieve reliability/safety is very rarely a function of technology. It's a function of how you do things. As are loss functions. (I'm an M.D., amateur physicist and a professional large systems designer, so I hear you on the MTBF stuff, but you are misassigning causes/effects and blaming the wrong thing). I suggest you reevaluate Deming because the very message of Out of the Crisis is what this whole little give-n-take is about..
Back to the top.
It's invisible to author of the article in question. He doesn't see the dynamics by which this comes about routinely in Western civ. And it has nothing to do with smart people not being able to make solutions. Repetitive of me , I know.
It has to do with process and incentive. Nothing else. Technical breakthroughs, admittedly usually come first, but isn't it interesting that the opposite occurs when process and incentive are properly aligned? (Concorde, 777, optical transistor, lcd panels, etc)
Within 24 months my organization will deploy an immunotherapy based on what was previously a dangerous and unreliable procedure which had as much a chance of curing you as killing you from a bleedout. We have shrunk it by a factor of 1000 and removed the bleedout danger. We did this by normal application of process and incentive and we have done it for under $5 million. (Our chief techie, amusingly, isn't even a biologist but is a materials engineer trained exactly as I have outlined.) It could have been done by the similar tech in 1987 when this therapy was first identified. It wasn't because the FDA/NIH removed all the incentives and screwed the process. (Conversely, we ignored the traditional path, didn't goto a major medical center, went to the vets who move 10 times faster, and only now will we apply with the FDA, and then not on their terms).
Well, back to the grindstone. Dynamics outweigh statics folks. 5 years from now, spacecraft reliability will where planes where in 1950.
"spacecraft reliability will (be) where planes were in 1950."
That's a pretty bold assessment. Also, are you aware that Airplane fatalities were substantially higher in the 70's once more people could afford to fly? Observe these statistics-
So, do you honestly think that safety will be designed in to space travel in a way that would be substantially "safer" than NASA designs? Because until the government enforced specific safety regulations in the late 70's, air travel was not on the path towards more safety and security. The market won't dictate safety, only what the value of space travel is. And someone strapping themselves on to a version of spaceshipone most likely has already examined the substantial risks associated with such a dangerous activity and ignored it. The market will not force better safety than NASA in to private space flights. It would be reasonable to assume that until enough people die and the government is forced to regulate the industry will safety even BE an issue.
Just my two cents: I think both Randall and brainpk may be right. What it seems to me is being left out is cost. It may be possible to achieve the required levels of safety with current rocket technology, but in doing so you've added so much overhead that it's no longer cost effective.
The classic example for me is nuclear energy (fission, not fusion). As most of us know it was once the "too cheap to meter" technology, and if there where no safety considerations, it probably would be. However, once you added safety in to the design, the cost went up to where it is today. I know some will point out that it's still pretty damn cheap and that, at least in the U.S., lack of enthusiasm is more a political issue (I would agree). Nevertheless, it's hardly "too cheap to meter."
BTW, as an engineer myself, and not using anything other than my instincts, it seems that any design which requires you to 1) sit on top of enough fuel to get you to orbit, and 2) burn all that fuel in just several minutes, is going to be inherently unsafe. I suspect that only a technology that slows the whole thing down will be safer. Space elevators would be ideal, but I don't think that's going to happen anytime soon either. Even the old idea of a space plane which takes maybe one to two hours to get to orbit would probably be safer. I could be wrong. Math anyone?
Airplanes were more reliable in the 1970s than in the 1950s. That greater number of fatalities is a function of a hugely greater number of flights.
Concorde was funded by the British and French governments. We don't really know how safe it is compared to conventional jumbo jets because it didn't fly as many total flights as conventional jets. Still, it suffered at least one crash that I can recall and that crash was due to tires throwing something up into the engine if memory serves.
Interesting. Such a static view. That's not how it really works. Failure to achieve reliability/safety is very rarely a function of technology.
Rarely does not equal never. There are certainly cases where engineers simply do not know how to make something safer or more reliable. Or they know how to do it but it costs too much. Yes, process technology can be improved. But it is far easier to improve process technology when you are making large quantities of something and therefore have sample sizes large enough to do statistical analysis and to refine the process and the design iteratively. For something like a rocket where at most dozens are made per year and in different designs for different mission profiles then, no, it is not easy to improve process technology or design. Many of the failure modes never come up and bite you.
Back when I was working at GM Delco I once accidentally got on a mailing list for a Chevy Nova plant's problem reports. It was pretty interesting to see what the problems were like. From one day to the next they'd find and fix some new problem that popped up with, say, window sealant or rubber tube quality from some supplier. They were pushing back in the supply chain get process improvements. It was working. But they were able to do this because of the volumes involved. Also, a car is orders of magnitude more simple to design than a rocket. Cars do not push materials technologies or design methodologies like a rocket can.
Then you say:
It's invisible to author of the article in question. He doesn't see the dynamics by which this comes about routinely in Western civ.
Alex is a free market libertarian economist who has a deep understanding of the dynamic processes that run Western civilization. I've met him and read lots of his writings. You ought to read his blog.
As for your safer immunotherapy: You probably couldn't have developed it in the 1960s because the supporting technologies were not available. That is where we are with rockets.
Yes, we need materials science advances to enable space elevator construction. Yes, that would probably make travel into space orders of magnitude safer.
Scramjets to orbit: There are materials and mathematical modelling problems in designing them. We simply do not know enough to do it.
As for the safety of nuclear power plants: More conventional refinement of light water reactor designs have made them safer. If Westinghouse ever builds an AP1100 then it will be much safer than the reactors currently operating because it is designed with fewer parts to fail and more passive elements. However, the next big step forward for both increased safety and lower cost will be the pebble bed modular reactors. They move much farther in the direction of designs that are inherently safer and more passive. Failures in PBMR can't cause a melt-down. But we do not have a feasible path to pursue to accomplish the same result with space launch. Our basic starting technologies and science aren't close to what we'd need to design an orders-of-magnitude safer space launch vehicle.
I understand that the planes were more reliable in the 70's, as well as the fact that there were simnply more of them, but most of the data I have seen shows that the downturn in fatalities in the 80's was primarily due to government safety requirements and regulating air travel in general moreso than the improvements on airplanes overall reliability. Analogous to the government requirements of seat belts in autos, which drastically lowered auto fatalities. The reason I make this point is that today, there are no current "safety requirements" for sub-orbital air travel beyond the regular requirements for private flight. When (and if I might add) sub-orbit travel becomes a reality in the marketplace, the first time someone dies from the spaceship who isn't on it (say an unfortunate victim on the ground) the government will be all over it and you can garauntee that unless requirements are met there will be no market for it in this country. Liability requirements will see to that...
If you want to make a comparison of airplane saftey for government controled to private enterprise, try looking up data on the saftey records of the US to the USSR during the early years of aviation. Though I am not sure where you would get records on the Soviets.
Anyways in my own biased opinion, the biggest difference between the slow progess in saftey (and other things) for spacecraft can be seen by the number of generations in designs for spacecraft compared to airplanes. For instance, the space shuttle main engines are arguably the second or third generation design for hydrogen/oxygen motors that produce thrust over 100,000kN and used in manned spaceflight. The first engine was the J-2 make for one of the upper stages of the Saturn V and was developed in the 1960's. Then came the Shuttle main engines which were developed in the 70's (there was a prototype made so you could say second or third generation, but I don't count it since it was never actually used in operations). After that the Soviets developed the RD-120 (their copy of the SSME) right after the space shuttle main engines were put into flight in the 80's. So after 40+ years we are still useing a second generation design to fly the space shuttle an you wonder why it has reliability problems.
Compare all of this now with the number of engines designed for aircraft over the first 40+ years of flight. You may notice a trend here. Other systems follow a similar pattern. The heat shield on the shuttle that can be argued to be a first generation 'reusable' design to be used on an operating vehicle. The second (or third if you count the Buran) generation reusable heatshield was going to be tested on the X-33, but that never flew.
Gus is right. Practical high-volume manned orbital flight will come only after many generations of design. I, too, don't think the history of aviation from 1903 to 1938 is the right analogy to the last 40-odd years of manned space flight. After all, 15 years after Kitty Hawk, we could build, among other things, Spad fighters and Gotha bombers; 35 years later, Spitfires and Boeing Clippers, with jets and helicopters to come in 1939. Aviation history types can name other record-setting or influential aircraft to fill the years in between. But what is the equivalent progress with _reusable_manned_spacecraft designs? By 1938 we understood the engineering well enough that aircraft had settled on a general form, i.e. single set of wings amidship, tail surfaces behind, retractable or fixed landing gear, engine in nose or wings, driving propellers. Do we really understand reusable-vehicular spaceflight to the point that we've settled upon an equivalent general form for spacecraft? Just to make sense of the 26 X-Prize contestants, MSNBC had to organize them into 4 flight schemes (http://www.msnbc.msn.com/id/5235594/).
Randall is right that 1938 aviation was safer than spaceflight is today. With two flight stages -- ground to air, then air to space (orbit), and back -- to handle, this is a much tougher technical challenge, and the necessary technical advances may not come as rapidly as with atmospheric flight. Alex basically says that rocket-propelled spaceflight is too dangerous to be safely carrying 100,000 space tourists a year in just 10-12 years time. However, airline industry data suggests that we were flying 100,000 passengers a year, 10-12 years after Kitty Hawk, and that aviation passengers of 1938 may have faced the 1-in-10000 chance of dying. (Randall's data plus industry data at http://www.airlines.org/econ/d.aspx?nid=1032). I would guess enough people thought the novelty or the time-savings was worth the cost and risk.
I do agree that it is unclear if cheap space tourism alone will draw enough (100K/yr) customers, particulary the "fat guys with cameras", given the 1-in-10,000 risk of death. Those who liken this to high-risk adventure travel may be right about Virgin's space tourism prospects.
Hmm, it seems I can't do arithmetic properly. In my previous post, I wrote that 10-12 years after Kitty Hawk, 100,000 passengers a year were flying on airlines. The data shows the time to achieve that level of traffic actually took about 26 years.
Gus is on the right track.
It isn't especially useful to compare the development of aircraft vs. spacecraft. In the early days of aircraft it was a highly entrepreneurial market and anyone could jump in. It took the government many years to get deeply involved in regulating the development and use of civilian aircraft. Spacecraft, on the other hand, have been under the government's thumb almost from the beginning. Almost every aspect of how access to orbit has been handled has served to keep private enterprise at arms length. The numerous companies now seeking to create practical vehicles for bringing humans and small payloads to orbits should have existed decades ago and likely would if NASA hadn't made it their mission to create the perception that getting into space was something only they could do.
The market for orbital access has been artificially retard for most of our lifetimes. It is only now that it is starting to take the form it should have always had.
Doesn't this erudite and intelligent discussion of safety miss the whole point? The cutting edge attraction of space tourism is exactly like that of guided climbs to the top of Mount Everest, the whole point is that it is dangerous! If it wasn't dangerous there'd be no point in doing it. The only people who can afford it can buy every safe experience on earth, what they want now is the thrill and there's no thrill, kudos or sense of achievement if there isn't a small chance of death. Everyone always thinks it's going to be the other guy who doesn't come back and no matter how many non expert, or expert, climbers die on Everest or K2, more people climb them every year.
A one in twenty chance of dying is acceptable to enough rich people to make this worthwhile as a commercial project. We're not at the stage of 747s landing 300 people in New York, this is much more akin to extreme mountaineering or drug taking or getting married in California. The people who make money these days are often thrill seekers or techno geeks - they'll know the risks and reach for the skies anyway. Anyone can go anywhere in the world for little money these days - your hairdresser goes absailing in Equador, that ten year old kid has climbed Ayres rock - the only other place to go is up and to keep going.
The safer we become as a culture the more obsessed with safety we become. This is why manned spaceflight becomes so incredibly expensive and why, nearly 40(!!!) years after landing on the moon we haven't been anywhere since. NASA should come clean. Be up front about the dangers and just get on with it and design a mission which isn't perfect or tested to death before it ever gets off the ground. Accept that, say, there's a 50% chance of a manned mission to Mars failing, of everyone dying. They'd still be swamped by applicants for astronauts. I'd do it and so would you.
Why do people like doing extreme sports? Scuba diving is dangerous, parachuting is dangerous, mankind is a risk taking species, it's what makes us so successfull. Commercial spacecraft will kill people, and they'll become safer and safer over time as a result. It's the fact that we launch so few manned craft that stops them becoming safer. There were hundreds of different aircraft flying millions of hours in 1938, it's less time than the number of different designs that are produced which spurs innovation and improvement.
The way forward is the much touted 'Mars Prize'. Offer ten billion for the first mission to land men on Mars for a certain amount of time, conduct a certain number of experiments, bring a certain amount of rock back and land safely. It's no risk for the government, it's high gain for society and it's the incentive that people need to make it happen. The only way the space 'race' is going to get going again otherwise is for China to land someone on the moon and wake the rest of us up.
Is pay per view reality TV going to fund the first mission to Mars maybe? I don't know. I do know that boys of my generation not only dreamed about going to Mars but actually thought they would. Now kids play video games set in fantasy pasts and the thrill of future technology goes no further than the next mini IPod. Perhaps it's part of a change in society. The brave men and women who gave their lives in pursuing their dreams of space flight were once heroes, but are now more likely to be looked upon as victims. It impoverishes and diminishes us all.