On June 21st of this year, with thousands in attendance in the small southern California desert town of Mojave, a sexagenarian test pilot performed a successful space flight of a private vehicle funded for less than thirty million dollars, by a co-founder of one of the world’s largest corporations.

Occurring about a year and a half after the loss of a second of NASA’s Space Shuttle orbiters, putting the nation’s vaunted manned space program on hiatus, it received a surprising amount of publicity. Our nation remains in a pause of our government-funded human space activities, other than a single American astronaut on the International Space Station, there only because he went up on a Russian vehicle. As these words are written, it remains unclear when the Shuttle will fly again, if ever.

One of the reasons that the story of SpaceShipOne was so compelling was that, a decade and a half after the fall of the Berlin wall, our space activities remain mired in institutions and thought patterns that are a four-plus-decade-old artifact of the Cold War.

The old, Cold-War space age with which most of us are familiar, and think the norm, is in fact, when viewed in the context of our nation’s total history, an anomaly. It is an historical contingency that could easily have manifested itself in a much different form had we not been locked at the time in a confrontation with totalitarian communism. Moreover, it is one at distinct odds with the traditional American values of individualism, freedom, and particularly free enterprise.

Despite the fact that SpaceShipOne only went into a suborbital trajectory (in other words, it didn’t have enough velocity to stay in space and orbit the earth), and spent just a few minutes in space and weightlessness, this contrast was still driven home quite graphically by that flight in Mojave in June. No one familiar with our more conventional, familiar manned space program could imagine NASA achieving the same feat for such seemingly little money.

Myths of the Old Space Age

Because we have seen few other examples of human space activities than those of NASA and the Soviet Union, a number of myths have developed and become entrenched in our minds since the dawn of the old space age, and they inhibit us from thinking about our new frontier in the more expansive ways necessary to open it up to all of us, rather than just a few astronauts. The word “myth” is used here in the cultural anthropological sense, as a broadly accepted belief by society, not necessarily in the sense that it is false. However, in fact, they are false, as I hope to demonstrate. In order to show this, and thus broaden our policy perspectives, this essay will examine each of them in turn. Let us first describe them:

Myth #1: Spaceflight, and particularly human spaceflight, is intrinsically expensive and dangerous, and only major governments have the resources to engage in it.

We are all familiar with news stories about the costs of NASA’s “reusable” space shuttles and space stations, and the billions of dollars spent not just in budget, but even in budget overruns.

Project Apollo cost twenty-five billion dollars in 1960s dollars, equivalent to well over a hundred billion dollars today.

The Space Shuttle cost several billion dollars to develop, and continues to cost billions of dollars per year to operate, even when, as now, it isn’t flying at all. Each flight costs between half a billion and a billion dollars, depending on how one does the accounting, not even including any amortization of the development costs.

The space station was originally projected to cost eight billion dollars to develop, a cost estimate that has ballooned to tens of billions since. Ignoring the fact that the original estimate didn’t include launch costs, and was based on a much smoother program with less continual redirection from Congress, even the original eight-billion-dollar estimate is a lot of money for a facility that supports just a few people, or at least, it would be, anywhere other than low earth orbit.

Given all of this empirical evidence, with no counter evidence to date, it’s not surprising that, like the famous four-hundred-dollar hammers that the Pentagon supposedly buys, we have come to expect human spaceflight, at least as performed by NASA, to be expensive. This brings us to the second myth, as a corollary to the first:

Myth #2: Spaceflight is intrinsically expensive because of the basic physics of it—it requires so much energy and power and propellants that we will never get the costs down with conventional rockets.

Conventional explanations of the high cost of space tend to be framed in terms of laws of physics. In order to get into orbit, a velocity of seventeen thousand miles an hour must be achieved. Rockets are the only way to propel vehicles to this velocity, because air-breathing propulsion systems, such as jet engines, don’t work in the vacuum of space. The amount of propellant required for a rocket is an exponential function of the velocity required (in other words, twice the velocity requires much more than twice the propellant, and it only goes up from there). Thus, to get a significant payload into space requires a rocket many times as large as the payload.

If the launch vehicle is expendable (as almost all vehicles in history have been, including the Saturn vehicle used successfully to send men to the moon thirty-five years ago), then the cost of getting a few thousand pounds into space becomes, almost literally, astronomical, because so much expensive hardware is only used once and then thrown away for each flight.

If the vehicle is reused, like the Shuttle, then it has to return to earth. Now a different kind of physics comes into play—the energy that was invested into the vehicle to get it into orbit must now be shed to allow it to slow down and land. The heat generated by this requires systems that can absorb and reflect it, and protect the vehicle from it. It was a failure of such a system that destroyed the Space Shuttle Columbia on entry on February 1, 2003.

In both cases—ascending and descending—the physical problem is one not just of the amount of energy involved (which is what many claim) but the rate at which that energy must be managed. The energy must be added very quickly for launch, and removed just as quickly for entry, so it’s really a matter of power—the rate of energy change.

All of these factors, so the argument goes, require complex systems, with many components, each of which must be reliable in order to make the overall system reliable, and all of this costs money.

Now compound this with the additional requirements for “human rating” such a system. Humans are priceless, and as important as multi-million-dollar satellites and launches are, putting human lives at risk requires even more reliability and safety. In addition, humans require life support—pressurized volume, breathable air, food and water—all of which adds even more weight to the needed payload. This in turn increases costs even further.

Space access is expensive because the fundamentals of physics make it so, and short of some new technology, like space elevators, it will remain so for the foreseeable future.

It’s a seductive argument, and a plausible one. It leads in turn to the next three myths:

Myth #3: Because manned spaceflight is so expensive, only major governments can afford to do it.

Until a couple months ago, the only existence proofs that humans could be sent into space were via massive government programs—the US and the Soviet Union (now Russia) in the early decades and, more recently, China. Thus, the notion that any entity lacking the resources of a major world power could accomplish such a feat has been met with almost universal incredulity. For those attempting to raise investment capital to do such a thing, this widespread belief, as well as that of Myth #1 (space is expensive), becomes a self-fulfilling prophecy, because no investor will be willing to put forth the funds for something that is perceived to be impossible. The other difficulty in raising funds comes from the next myth:

Myth #4: There is no use, or few uses, for humans in space that can justify their expense.

Note that the argument isn’t that there is no use for humans in space at all—the myth depends on the fact that it’s expensive to put people there (Myth #1—space is expensive).

Yes, the Apollo astronauts did good science, and astronauts have repaired satellites on orbit, but the benefits didn’t justify the costs, when it could have been done with robots, which are lighter and don’t require life support. This is particularly the case if the money being spent on expensive manned space systems had instead been used to develop smarter, more advanced robots.

In response to the argument that there are other reasons to have people in space than science and maintenance, that there are many people who might like to simply go for the experience, or even to build a new life off planet for their own reasons, and that these are purposes for which robots are...superfluous, we come to another myth, a consequence of all of the previous ones:

Myth #5: Public space travel, if it ever occurs, will only occur after needed technological breakthroughs, as a result of continuing government investment in NASA’s (and perhaps the Department of Defense’s) space programs, dramatically bring down the costs, and improving the reliability and safety.

This is certainly the conventional wisdom, and has been as long as I’ve been in the aerospace industry (which is a depressingly long time). If you buy into myths #1-4, it follows almost inexorably.

The argument here is that while a couple people have paid twenty million dollars for a spaceflight to the ISS on Russian hardware, the market at that price is not very big. We won’t be able to get prices lower without an investment in new launch technologies (that probably don’t involve crude rockets), and the market is too small for any private investor to make such an investment, given the uncertainty in the shape of the price-demand elasticity curve (i.e., the rate at which the market increases with a decrease in price), because such an investment will involve billions.

Only governments can afford to make that investment, and until they do (perhaps in order to reduce their own costs of operation for civil and military space), we will be stuck with high launch costs. However, once costs have been reduced with appropriate government investment, if that occurs, perhaps decades in the future, then there’s a possibility that those government systems can be converted to private ones carrying paying passengers, or that commercial derivatives can be developed.

And finally, here’s one that, sadly, both NASA and the Aldridge Commission (and the administration in general) seem to have accepted:

Myth #6: We don’t have the technology to build reusable launch vehicles, or if we do build them, they cannot fulfill their promise of low launch costs.

We thought in the 1970s that reusability was the key to cost reduction, but we invested billions of dollars in the Space Shuttle, and it has turned out to cost more per pound of payload delivered than the Saturn V that it replaced, and it has killed fourteen astronauts. In the 1990s, we instituted a reusable experimental program called X-33, but after spending a billion dollars on it, it never even flew, and it now sits up in a hangar in the Mojave Desert along with a launch pad that it never used. The head of NASA’s Marshall Space Flight Center declared a couple years ago that X-33 proved that we just don’t have the technology to build a reusable vehicle, and after all, if he doesn’t know, who does?

It’s time to give up this dream until the technology has come along further, and return for now to the capsules on expendable launch systems that successfully got us to the moon in the 1960s.

There are other myths of the old space age (e.g., space is for science only, we can’t afford large space programs without international cooperation, etc.) but the six I’ve listed are the ones primarily responsible for holding back progress. My intent was to state them, in good faith, as I believe those who hold them would—I’m attempting to avoid the creation of strawman arguments, easily knocked down, that no one actually believes. I hope that I’ve succeeded.

Nonetheless, they are refutable, and I claim that they are all false beliefs, as I hope to show next.

Myths Debunked

Let’s start with myth numbers 1 and 2, the notions that space is intrinsically expensive, and that this is because the physics demands it—that after almost half a century, rocket technology is completely mature, and has reduced costs just about as far as they can go. These are the linchpins for the others.

In order to show this to be mistaken, we have to both show why physics isn’t necessarily the problem (thus exonerating Isaac Newton), and identify the true reasons for the high cost of spaceflight (and particularly manned spaceflight), thus providing some guidance as to how to actually achieve low launch costs.

I’ve already said that the argumentum ab physics sounds plausible, so what’s wrong with it? Before we get into that, let’s consider some alternative explanations for high launch costs.

Back in the 1980s and early 1990s, as an employee of a major government aerospace contractor, I participated in and managed several studies relating to future launch systems. These were so-called “space transportation architecture” studies in which alternate conceptual launch systems were evaluated and compared to each other, in order to provide some guidance to decision makers at the Air Force and NASA as to what future investments to make in launch vehicles and technology.

These studies considered a wide variety of vehicle types—reusable, expendable, single and multiple stage, various propellant combinations, air breathing, rocket, horizontal takeoff and landing, vertical takeoff and landing, etc.—the entire range of conceivable ways of getting crew and cargo into earth orbit and (when necessary) back using semi-conventional aerospace vehicles. They also considered a range of potential “mission models,” scenarios of types, mass and volumes of payloads, as a function of time out over the next few decades. Some of the models were minimal—no commercial activity, little or no growth in NASA/DoD space budgets—while others were expansive—major new civil space initiatives, including crewed lunar and Mars missions, and large-scale commercial activity (though no space hotels or tourism), with others falling in between these two extremes.

As we looked at all the combinations of architectures and models, we discovered something interesting (and something that was rarely emphasized in the reports to the customer). While some vehicle design concepts were clearly better than others, for the low-activity scenarios, they were all extremely expensive per flight, and for the high-activity scenarios, they were all much less expensive. In fact, we considered the Shuttle as a reference, and developed a notional architecture that had sufficient facilities and vehicles for a hundred Shuttle flights per year (not as ridiculous as it sounds today—the original Shuttle was expected to fly over once a week, until its design was compromised by inadequate development funding in the seventies). Surprisingly, while it wasn’t the cheapest system by any means, its estimated costs were nonetheless dramatically lower per flight than the actual Shuttle costs at the flight rate of the time (less than ten per year).

For any given mission model, the range of life-cycle costs (including development costs) for candidate launch systems wasn’t that wide. My recollection is that the difference between the most and least expensive might have been a factor of three.

But the difference in cost, on a unit basis (that is, cost per flight, or cost per pound) varied dramatically for different mission models. In some cases it could differ by more than an order of magnitude (a factor of ten) for the same vehicle, with the only difference being the level of activity.

What this means is that the theoretical best vehicle concept, flown rarely, will be unaffordable to fly. A mediocre design, flown often, will beat it in cost per flight. In other words, the vehicle design had what a mathematician would call a second-order effect on costs, but the flight rate had a first-order effect. How much you used the launch system was much more important than what kind of propellant it used, or how many stages it had, or whether it took off or landed horizontally or vertically, or any other design choice. This, to me, was the key insight from all of those studies, and it’s one that remains true to this day. This doesn’t mean, of course, that we should ignore vehicle design, but it does mean that we need to pay much more attention to markets.

The reason for this is obvious, in retrospect.

Consider a mundane example. Do a thought experiment, in which Boeing developed the 747, spending several billions of dollars in doing so. But instead of building hundreds, and flying each of them daily (as is the case), they only built five, and only flew each one once per year. The design is exactly the same—the only difference is the number of vehicles built and the flight rate.

Just to use some reasonable but simplified numbers, imagine a ten-billion-dollar development cost, including the production of the five airplanes. Let’s say that Boeing didn’t make any profit, but sold the five airplanes to American Airlines for two billion dollars apiece. Now, if American can borrow money at less than ten-percent interest, it has annual costs in aircraft payments of over two hundred million dollars per year, including paying down the principal at some rate, for each airplane. Even if they have absolutely no other expenses (fuel, pilots, flight attendants, marketing, ticket agents, etc), if the aircraft has four hundred seats they would have to charge half a million dollars per ticket just to cover their loan for the aircraft purchase.

Note that the physics was exactly the same, and the aircraft design was exactly the same, as that for the Boeing 747 that is built in quantities of hundreds and flown every day. In one case (the real world), the ticket price is a thousand dollars. In the other case, it’s five hundred times that, and we didn’t even consider the fuel costs, or the army of support personnel necessary to support a fleet of aircraft, whether it’s five or five hundred. In other words, if we ran the aviation industry in the same manner that we run the Space Shuttle program, it would be almost as costly.

Here’s another example, from real life, and recently. The Air Force and its launch contractors invested billions of dollars in the Evolved Expendable Launch Vehicle (EELV) program in the 1990s, with the goals of improving reliability, and reducing costs of its existing Delta and Atlas launch vehicles. The specific goal was to reduce their operating costs by twenty five percent. In other words, they determined that by coming up with a new vehicle design, the best they could do was reduce costs by a quarter. Some might interpret that as vindication for the argumentum ab physics—a sign that the launch industry is indeed mature and that very little marginal improvement can be expected from conventional rockets in the future.

But consider—the cost estimate was based on a robust (for launch vehicles, at any rate) flight rate from both Air Force and commercial payloads, including geostationary communications satellites. When the bubble popped in the year 2000, one of the many casualties was the communications satellite market. Boeing has dropped out of the commercial launch market as a result, and the total expected flight rate for the Boeing Delta and Lockheed-Martin Atlas EELVs has plunged, resulting in a per-flight cost rise of up to fifty percent. That means that a simple change in market had a much larger effect on launch costs than the billions of dollars spent on redesigning the launchers.

While the lack of economies of scale is an important, and perhaps the dominant reason that launch costs are high, it isn’t the only one. Another is that, as a result of our space activities being dominated by government contracts from the beginning, with purposes for which cost was no object (defending the nation with ballistic missiles, and beating the Soviets to the moon), we have developed an industry culture in which high costs are expected and accepted. An additional problem is that, despite all the lofty rhetoric about science, and exploration, the civil space program is supported in Congress largely because it provides jobs in key congressional districts, whose representatives get assigned to committees overseeing its budget.

When contractors have contracts that reimburse them for all costs, plus a percentage of that as profit, and when creation of constituent jobs, rather than conservation of taxpayers’ funds, is the congressional priority, it’s not surprising that government space programs are expensive.

These aren’t, of course, the only reasons that launch costs are high. Physics is a problem, and getting things into space is a technical challenge. But it’s not enough to justify the existing high costs of access to orbit. Depending on which vehicles one considers, and how one assesses cost versus price, there is somewhere between two and three orders of magnitude difference between unit costs for air travel versus those for space flight reasonable estimate is that the technological issues account for perhaps one order of magnitude of the cost difference between air travel and space travel. The other one or two orders of magnitude are almost certainly due to the institutional practices (including the fact that the vehicles are usually thrown away) and lack of economies of scale.

One hint that this might be the case is the relative allocation of costs in the airline industry versus those for space launch. Ultimately, the cost of fuel places a floor on the cost of an airplane ticket. No matter how productive the airline staff becomes, and no matter how cheap the aircraft themselves become, you’ll never get the average price of a ticket below the cost of the energy necessary to deliver a passenger from point A to point B. And indeed, when we break down the costs, it turns out that fuel costs are the second highest item, after labor, and typically account for somewhere between twenty and forty percent of the total.

A naïve observer, seeing the size of the propellant tanks on a rocket, might assume that propellant costs for a launch are tremendous. But in fact, propellant costs are an insignificant proportion of the total costs. As an example, the new Boeing Delta 4 launch vehicle requires over half a million pounds of liquid oxygen and liquid hydrogen to get its payload to orbit. The costs of these propellants is far less than a dollar a pound (the price of liquid oxygen is actually comparable to that of milk). But even if we grossly overstate their costs, and say they’re a dollar a pound, this is still only half a million dollars for the propellants. This for a launcher that costs (or at least is priced) on the order of a hundred million dollars per flight. So for launch, as opposed to an airliner, fuel costs are less than one percent of the total costs. This would suggest that, contrary to conventional wisdom, launch technology and infrastructure using rockets isn’t even approaching maturity, let alone obtained that state.

It would also suggest that, if we could get to the point at which propellant costs do become significant, comparable to airline tickets, that launch costs would be much lower. For the Delta example, a half million pounds of propellant might deliver about fifteen thousand pounds into low earth orbit (it’s hard to calculate with certainty—Boeing only seems to provide public performance data to the more demanding geosynchronous transfer orbit). If propellant costs were thirty percent of the total costs, that means that the total launch cost would be about one and a half million dollars. That works out to only a hundred dollars per pound (two orders of magnitude below the typical costs today), which is an indicator of where launch costs could go, even using rocket technology, if we operated more like airlines.

Of course, the biggest and simplest reason that a Delta launch is so expensive is that we throw a very expensive vehicle away each time we use it. This gets us to Myth #6—that reusable vehicles are not the key to reducing launch costs, and that they can’t even be built. In the late sixties, it was common wisdom, and common sense, that we couldn’t afford to continue to, in Arthur C. Clarke’s words, carpet the bottom of the Atlantic with spent hardware—that we needed a reusable launch system. This was the thinking that led to the development of the Space Shuttle, and it should be obvious why.

Let’s go back to our Boeing 747 thought experiment except that, this time, instead of only flying the airplane once a year, and then reusing it again the next year (the Shuttle analogy), we burn it at the end of the runway each time we land (the expendable analogy). Now we have to buy a new airplane each time we fly. The marginal cost is much less than the two billion dollars than each of the first five cost us, because we’ll be amortizing the development and fixed costs over a larger number of units (a 747 goes for something like a couple hundred million dollars at current production rates). So now, instead of paying a half a million per ticket to pay for the borrowing costs for a reusable, but seldom-used airplane, we’re paying the same amount to buy a new airplane for every flight. And note that economies of scale don’t help much here. In the first case, where we reuse the airplane, we can dramatically reduce ticket prices by simply increasing the flight rate. This won’t work for the expendable airplane case.

By now, my hope is that it’s obvious why space launch is so expensive, and that it’s not because of the amount of energy required, or laws of physics in general. It’s because of the ways that we’ve chosen to do it, and the fact that we do so little of it. In the case of the expendables, it’s because we throw expensive hardware away with every flight, and in the case of reusables, it’s because we don’t reuse them very much. “Keep the wheels in the wheel well” is just as important an operating principle for low operating costs of a spaceline as it is for an airline.

As described in Myth #6, people illogically conclude from the failures of the Shuttle, X-33 and X-34 programs that we were mistaken in our beliefs of the late sixties—that reusables in fact don’t make sense. This is the fallacy of hasty generalization. One cannot draw grand conclusions about a general class from one or two data points.

Particularly absurd (and tremendously damaging to public and investor perception) was Marshall Space Flight Center’s director Art Stephenson’s statement in March of 2001, that “We have gained a tremendous amount of knowledge from these X-programs, but one of the things we have learned is that our technology has not yet advanced to the point that we can successfully develop a new reusable launch vehicle that substantially improves safety, reliability and affordability."

X-33 was a Marshall program in the 1990s to build a test vehicle for demonstration of technologies for a single-stage-to-orbit launcher (i.e., it wouldn’t drop any stages on its way to orbit—the entire vehicle would go all the way into space, and return). It wasn’t designed to go to orbit, but it was a subscale demonstrator that would supposedly demonstrate that a larger orbital version (called “Venture Star” by the winning contractor, Lockheed Martin) could be built commercially. Because it incorporated several different high-risk technologies in a single vehicle, it was a very risky program, and in fact was chosen for that reason—NASA wanted to “push the technological envelope,” so to speak. A key element of the program was that part of the proposal evaluation would be based on a “business plan” that had to be submitted, which purported to show how the winning contractor would go forward with development with its own (or at least private, not taxpayers’) money.

Unfortunately, Lockheed-Martin had almost no commercial experience since their last ill-fated attempt at an airliner, the L-1011 Tri-Star, in the 1970s—they were (and still are) almost exclusively a government contractor. Their business plan, such as it was, relied on taking over the existing geostationary communications satellite market and the Space Shuttle missions (primarily support to the International Space Station), but didn’t propose to open up any significant new markets (necessary to get the flight rate necessary to justify the development of a reusable vehicle), and it appeared at the time to be unrealistic.

However, one cynical view is that simply by winning the contract, Lockheed-Martin achieved their strategic business objective, since either they would succeed, in which case they’d take the business away from their competitors, or they’d fail, in which case, for relatively little investment on their part, they’d prevent the development of a competitor to their own existing launch vehicles, the Atlas, Titan and their half of the Shuttle (through their subsidiary United Space Alliance, co-owned with Boeing, which operates the Shuttle for NASA). Keeping existing cash cows alive could well have been worth the few tens of millions (of cash profit—not contributions in kind or government-reimbursed Independent Research and Development funds) they invested in the program, considering that they were keeping many hundreds of millions away from their competition (at the time, Rockwell and McDonnell Douglas, now both part of Boeing). It apparently didn’t occur to NASA that, by selecting an incumbent launch vehicle provider, there were intrinsic incentives for program failure. Nor did it occur them, more generally, that one shouldn’t seek innovation from an entity that’s benefiting from the status quo.

In any event, it never flew and was cancelled in the early part of this decade after spending more than a billion dollars on it. It’s important to understand that the X-33 was a demonstrator for building a single-stage launch system, but that’s not necessary for reuse, and X-33 itself wasn’t even necessary to build a single-stage system unless one chooses to do so with the particular technologies that it was supposed to demonstrate.

So in fact, all that X-33 (and X-34, another failed Marshall program) proved was that we didn’t have the technologies to build an X-33 and an X-34, few, if any, of which were essential to building a generic reusable launch system. (Some, less charitable than me, might say that it also proved Marshall’s ineptness at managing the development of experimental flight vehicles, and recognizing bogus business plans).

To say that the failures of Shuttle and X-33 prove that reusable launch systems cannot be built is logically equivalent to saying that all numbers are prime because the numbers “two” and “three” are. This is particularly the case when one considers that we’ve never actually even built a reusable launcher. It’s easy to forget that the Shuttle is only partially reusable. The external propellant tank is thrown away each time, and it costs tens of millions of dollars. The Solid Rocket Boosters (SRBs) are rebuilt each time, but only the casings are reused, after being reassembled into essentially new SRBs. Only the orbiter itself can be said to be reusable, and even that requires a great deal of between-flight maintenance, due to cost cutting during development.

In fact, the Shuttle actually demonstrates that a reusable vehicle can be built, and be quite reliable, if we consider only the orbiter. In neither the case of the Challenger or the Columbia was the orbiter the cause of the disaster. Challenger was destroyed as a result of the failure of an (expendable—they’re replaced each flight) O-ring in the rebuilt SRB. While the Columbia’s thermal protection system failed to protect her, this was a result of an external event—a foam strike from the expendable external tank. It makes no more sense to blame the orbiter for this accident than it would to blame an airplane if it’s hit by a piece of debris for which it wasn’t designed. Get rid of the expendable components, and neither event would have occurred.

When it comes to reliability, there’s one more important point to be made. Again, it’s conventional wisdom that rocket launches cannot be reliable, with the physics once again to blame. Even the new EELVs, supposedly the epitome of the state of the art in rocketry, only claim a 98% reliability. This would be appalling in any other form of transportation. But the reasons that rockets aren’t reliable aren’t due to the Evil Physics—they are the same reasons they cost so much. We don’t fly them enough, and we throw them away.

As an example, consider the Japanese. While their cars, in the sixties and seventies, were originally considered a joke—motorcycles on four wheels, they imported American quality control techniques, in the 1980s and 1990s they’ve become renowned for the quality and reliability of the automobiles.

Unfortunately, in developing their own space capabilities, the Japanese instead took a cue from our own failed space activities, having no successful ones to emulate. Like NASA and, for the most part, the Air Force, they deluded themselves that affordable and reliable launchers could be built by souping up ballistic missiles and flying them a few times a year. As a result, they too have unreliable launch systems, which resulted last year in the loss of critical military surveillance satellites.

The biggest difference between Japan's auto industry and Japan's rocket industry is not the almost unfathomable power that the rockets put out, or the harsh environment in which they operate, or their high cost per rocket. The biggest difference is that, as noted above, they built millions of cars, and they've built, at most, dozens of rockets.

There's an old aphorism that "quantity has a quality all its own." For this particular case, there's a reverse corollary--quality requires a quantity all its own. Statistical quality control is very useful when running a million cans of beans, or a million Honda Accords off a production line.

However, it's meaningless when only building a few of something, and only using (and in this case, expending) them a few times a year. There's almost no measurable learning curve, and because they're expendable, making each flight a first flight, there's no opportunity for the traditional "shakedown cruise." Infant mortality is high, and so, when we lose the baby, we also lose the bath water, the bathtub, the bathroom, and the house that contains them all.

So, if my thesis is valid, then it means that the key to low-cost and reliable launch is 1) stop throwing the launch vehicles away, in whole or in part, and 2) fly them a lot. To paraphrase the old 1992 campaign slogan, “It’s the flight rate, stupid.” Which in turn means that “It’s the market, stupid.”

And if that’s true, what are the implications for our space policy, and in particular, for the president’s new Vision for Space Exploration?

Failed Paradigms

It is clear, as already noted, that the administration doesn’t believe that truly reusable launch vehicles (an ugly and misleading phrase that I’ll henceforth replace with, as suggested by Mitchell Burnside Clapp of Pioneer Rocketplane, “space transports”) are technically feasible. NASA ended the Space Launch Initiative last year, a program whose goal was to prove out the technologies needed for space transports, converting its funding first to the Orbital Space Plane program, and then ending it entirely.

Now this wasn’t necessarily a bad thing to do, because, as noted, technology per se is not the issue. But as is often the case, when government occasionally does the right thing, it’s often, even usually, for the wrong reason, and this was no exception. They didn’t cancel it because it was the wrong way of achieving the goal (which it was). They did it because they don’t believe that the goal of significantly reducing launch costs is achievable. They’ve thrown in the towel, and resigned themselves, and perhaps consigned the nation, to high launch costs with low reliability for the foreseeable future.

However, they now have a policy disconnect, because a key element of the president’s new policy is that human exploration be “affordable and sustainable.” In returning to the sixties, Apollo-like philosophy of sending humans to and from orbit in capsules on expendable launchers (which is the essence of the new “Crew Exploration Vehicle” or CEV program), they choose an architecture that will be exactly the opposite. Apollo was demonstrably neither affordable or sustainable, since we didn’t sustain it (largely because of the perception that we couldn’t afford it).

While the report of the Aldridge Commission on the new vision, released in June, had some good recommendations in it, it also had a few potentially disastrous ones. Perhaps the most damaging statement in it was to declare heavy-lift launch systems to be an “enabling technology” for carrying out the vision. This is a phrase of art in the engineering world meaning that, absent such a technology, the goal is unachievable. The commission is claiming that we cannot send humans beyond low earth orbit without a much larger launch vehicle than anything existing. If they had used the phrase “enhancing technology,” meaning that it’s not an absolute necessity, but that it makes things easier to do, I’d have less complaint, but as they’ve stated it, it commits us to an expensive development of a new launch system, that shows no promise of actually reducing costs. Moreover, it commits us to an approach to exploration that, like Apollo, is not affordable or sustainable.

Recall my thesis that, in order to reduce costs, we must dramatically increase flight rate. For any given level of activity, the larger the vehicle, the lower the flight rate, because more payload can be put up on a given launch. At least initially, my sense of the president’s plan is that it doesn’t envision more than a flight or two to the moon per year. So we will spend a lot of money up front, to build a vehicle that will seldom fly, at a high cost per flight. Some might argue that this simply means that we have an opportunity to therefore fly more missions, taking advantage of our newfound capability. The problem is that increasing the number of flights of such an oversized vehicle (particularly if expendable, and no one has seriously proposed a reusable one) will quickly break the budget.

In addition, a heavy-lift vehicle for a single-mission launch carries a lot of precious eggs in a single basket. Since reliability comes from increased flight rate and experience as well, this means that we are placing a tremendous amount of risk on each of our missions.

There are two fundamental philosophical approaches to space operations, and there’s no intrinsic reason that either can’t work (at least in terms of being able to physically accomplish the mission—sustainability and affordability are a separate issue).

One is to launch everything required for a mission in a single launch. The other is to deliver things in pieces, and assemble them on orbit.

This was a debate that occurred during the early planning of Apollo. Those who advocated a single-launch approach won out, resulting in the successful landing of men on the moon within Kennedy’s deadline of “within this decade.” They won out for good reason, but it’s important to understand that reason (or those reasons), so we can understand why, forty years on, those reasons no longer apply.

The purpose of Apollo was not to build a sustainable capability to explore the solar system. It was, just as Kennedy said, to put a man on the moon and return him safely to earth, within the decade. Beating the Soviets to the moon was the primary goal, not science, not building infrastructure. In fact, the NASA administrator at the time, Jim Webb, was concerned that Apollo would result in a dead end, and tried to get Kennedy to fund a more flexible and ambitious program, at which point the president told him that he “wasn’t all that interested in space.”

Thus, every decision made in the Apollo program was made to meet the lunar landing goal, and everything else was subordinate to it. Wernher von Braun’s initial plan was a scenario called earth orbit rendezvous (EOR), requiring multiple launches whose payloads would be mated in orbit before heading off to the moon, but we had very little experience with rendezvous and docking at the time (before it had been proven with the Gemini program), and it was judged to be too risky. Thus we went with a plan that allowed the mission to be performed with a single launch of a huge launcher. But it wasn’t done because it was the cheapest, or because it was the best way of building a sustained capability or infrastructure. The choice was made because it was the most certain way to achieve the only real goal—to beat the Soviets to the moon—and cost was no object.

But that was then, and this is now. As President Bush said in his speech on January 14th of this year, in which he laid out the new Vision for Space Exploration, it’s not a race, but a journey. The goal this time is to build capability, and do it in a way that builds momentum, both politically and economically, so it doesn’t falter as we did after Apollo.

So how do we accomplish that?

First, we have to unburden ourselves of the confining myths of the old space age. Mark Twain wrote that a cat that sat once on a hot stove would never do that again. But it would never sit on a cold one, either.

We seem to be taking the same approach with space. We are convinced by the Shuttle experience that we cannot build affordable space transports. We are convinced by the space station experience that we must avoid on-orbit assembly. This is what Henry Spencer, redoubtable amateur space historian, has memorably called the “Wile E. Coyote” approach to engineering, after the hapless and unsuccessful hunter of the Roadrunner in the Warner Brothers cartoons. Not in the sense that we order everything from ACME, of course, but in that we try a particular technical approach, and when it fails spectacularly, we simply drop it and try something completely different, rather than examining what went wrong, and incrementally improving the concept.

Shuttle didn’t fail because it was reusable. It failed because it wasn’t reusable enough, and because it had too many conflicting requirements, and because it wasn’t funded properly during development, and because the true goal of the program was not to reduce launch costs, but rather to give NASA something to do after Apollo wound down and maintain jobs in politically important states and congressional districts.

Similarly, ISS didn’t fail because it had too much orbital assembly. It failed because we’ve never developed the necessary orbital assembly tools and capabilities, and because it, too, had too many conflicting and ever-changing politically imposed requirements, and because we didn’t have routine and affordable access to orbit, and because the goal of the program was not to build a space station, but rather to give NASA something to do after the Shuttle development wound down, and to justify the development of the Shuttle, and later, in the 1990s, to promote “international cooperation.”

And all of these causes had a more fundamental cause. Ultimately, the failure of both programs can be attributed to a failed paradigm—that of a beneficent and competent government agency as the trailblazer in the new frontier, complete with five- and ten-year plans. Of necessity, in response to the Soviet socialist state enterprise for space, we created one of our own, except it was based on a democratic political system instead of a totalitarian one. It worked fine for a short moon race, but we made the mistake of thinking, in utter defiance of our nation’s historic tradition of individualism and free enterprise, that this could be a successful model for the sustained opening of a new frontier.

The president’s new policy shows some burgeoning signs of recognizing this, but whether out of a failure to fully understand the problem, or out of practical political constraints, it doesn’t quite go all the way. It is encouraging that the president and the Aldridge Commission urged NASA to work with space entrepreneurs and private enterprise, to integrate them into the vision. But it still gives NASA the lead in developing a new vehicle for delivering humans to and from space, and doesn’t require them to purchase such services from the private sector.

But the planned “Crew Exploration Vehicle” (and a new heavy-lift vehicle, if NASA is allowed to develop one) will follow in the same failed path as the Shuttle and the station. Misreading the lessons of history, the new policy is born of a nostalgia for Apollo, the space program that was successful, so we return to what we mistakenly think made it successful—a capsule with an escape system on a large launcher. But it wasn’t the technical approach that made Apollo successful—it was the limited nature of the goal, and the national importance of achieving it. Apollo could have been successful with some other technical approach.

The issue is not the technical approach, but the programmatic approach or, at an even higher level, the fundamental philosophy. Because when a government takes an approach, it is an approach, not a variety of approaches. The study contracts are let, the contractors study and compete, the government evaluates, but ultimately, a single solution is chosen with a contractor to build it. There has been some talk of a “fly off” for the CEV program (something that might have mitigated the disaster that was X-33, had there been funds for it), in which two competing designs will actually fly to determine which is the best. But still, in the end, there will be only one. Just as, if we decide to build a heavy lifter, there will almost certainly be only one, since it will be enough of a challenge to get the funds for that, let alone two.

Biologists teach us that monocultures are fragile. They are subject to catastrophic failure (e.g., the Irish potato famine). This is just as true with technological monocultures, and we’ve seen it twice now in the last two decades, each time a Shuttle orbiter was lost, and our manned space program was put on a hiatus for years. We cannot reach our one and only space station right now, because we (that is, the US) had one and only one way of getting to it, and that means has been shut down since February 1, 2003, with no date certain for it to be once again available.

We understood this back in the old Space Transportation Architecture Studies when, in the wake of the loss of the Challenger, we were asked by the government customer to add “resiliency” to the list of desirable attributes for an architecture. But we’ve since apparently forgotten it, as the Air Force now talks about eliminating one of the EELVs (either Boeing’s Delta or Lockheed-Martin’s Atlas), because there’s not enough business to maintain both, and as the new space policy proposes a “Crew Exploration Vehicle” and a new heavy-lift vehicle. The same flawed thinking reared its ugly head during the Space Launch Initiative and “Next-Generation Launch System” programs, in which talk of the “Shuttle replacement” (singular) was prevalent.

In Virginia Postrel’s seminal book, The Future And Its Enemies, she talks about “stasists” versus “dynamists”—a different and more interesting way of looking at ideological divides than the traditional, but stale and useless labels of “liberal” versus “conservative.” To stasists, planning and order, and avoidance of change and particularly unplanned change, are the highest values. Dynamists are more interested in organic and emergent market-based solutions to problems—not as predictable, but ultimately more resilient and satisfactory to individuals.

While the state here has been growing at an uncomfortable rate to many for the past few decades, the tradition for the U. S. has been one of dynamism. But for the past forty-five years, largely because of the nature of its birth, nurtured in the fear and urgency of the Cold War, our national space policy has been one of stasism. And for all of its vision, the president’s new initiative remains at its heart a stasist one, though in its call for more participation from free enterprise, it contains the seeds for dynamism

What would a truly dynamist space civilization look like?

Imagine, instead of a launch of a few government employees every couple of months or so, daily trips into space of hundreds or thousands of private citizens. Some of them are off to do research at private orbital laboratories, some are heading to orbital resorts, yet others are boarding cruise liners for trips around the moon. Electric-propulsion tugs are hauling water up to a gateway as fuel for lunar landings. There are hotels in high inclination orbits for spectacular views of the earth, and vehicle assembly hangars in low inclination for departure to points beyond earth orbit. There are huge radio telescopes on the far side of the moon, protected from the incessant noise of our industrial planet, and at the poles are research facilities and tourist resorts, using the water ice hidden in the craters there.

The vast majority of the funding for this come from private expenditures, by people seeking their own pleasure and adventures off planet, and NASA has little involvement, other than to take advantage of the dramatic reductions in cost that it has created to do those things that only it can do, such as expeditions to the outer planets.

Is this a science-fiction fantasy, or economically and technologically realistic? How would we get there?

As Tom Wolfe entertainingly chronicled in his classic book The Right Stuff, over four decades ago, as Lyndon Johnson declared that our nation wouldn’t go to bed by the light of a communist moon, while the Germans, refugees from Hitler’s rocket program, were in Alabama developing the vehicles that would eventually take us there, there were rocket planes flying in the Mojave desert, released from B-52 bombers. They sundered the skies, testing, probing the upper reaches of the atmosphere and even temporarily leaving it. These were the first, tentative space vehicles, and had they not been interrupted by the urgency of beating the godless commies to the moon, their successors, of improved (and perhaps quite different, based on lessons learned) design, might have continued. The might have flown higher, and faster, and faster yet, until at last they flew fast enough to defy the gravity of the earth, falling at the same speed as the curving planetary surface fell beneath them, so they were in orbit.

That might have been another road to space—in the words of the poet Robert Frost, a path not taken—and one that might have provided a more measured, incremental, affordable and reliable approach, instead of one in which we decided to put small capsules on unreliable and expensive munitions, and hoped for the best. As with all alternate histories, we’ll never know, of course.

A New Path

But perhaps we saw the embryo of a new, dynamist space age, the one that was short circuited by the Cold War, in the clear blue skies over the Mojave Desert in June. SpaceShipOne was built in response to a private prize, a purse put up to urge private activities to seek the heavens, just as a private prize drew Charles Lindbergh across the Atlantic over seven decades ago. It’s the most successful of the contenders, but it’s by no means the only one, and the ten-million-dollar purse has generated many times that amount in efforts to win the prize. SpaceShipOne itself has reportedly cost more than the ten-million-dollar prize value, but no one complains. Contrast this to cost-plus-fixed-fee contracts to the traditional aerospace industry.

As these words are written, its backers have given their sixty-day notice of their intent to go for the prize, with the first of the two required flights to be attempted on September 29th. If all goes well, in early October of this year, the prize will have been won, and private astronauts will have flown into space on a private space transport twice within two weeks. The marginal cost of such a flight will be much less than a million dollars—a price within the range of many people in this wealthy country. But the important thing is that there won’t just be a winner, but someone who places, and someone else who shows. And the vehicle that wins the prize may not be the vehicle that taps the potential new markets. The Wrights, after all, were the first to master controlled flight, but they weren’t the ones to benefit the most financially from it.

But it’s reasonable to ask if there’s really a market, and how such vehicles, capable of only a tiny fraction of the energy needed to achieve orbit, can really open up the cosmos to humanity.

To address the second question first, it’s important to understand that the approach of the government to low-cost space flight has traditionally been to figure out how to achieve a given (high) performance level, and then figure out how to make it cheap. It has also been an approach, like that of the mythical birth of the Greek goddess Athena, of springing fully-grown from the brow of Zeus, after billions of dollars spent in analysis and test.

Saturn was not an incremental development—the first flight was all up. In fact, that was the innovation that Apollo program manager George Mueller brought to the program that allowed us to achieve Kennedy’s schedule-driven goal. But it wasn’t cheap. Similarly, other than drop tests of the orbiter to determine its flying capabilities, and an on-pad test firing of the main engines, the very first flight of the Space Shuttle in the spring of 1981 went all the way to orbit, with a crew of two. It wasn’t possible to fly a few feet off the pad, hover, and return, because its landing mode (horizontal) was different from its launch mode (vertical). Once those huge Solid Rocket Boosters light, there’s no way to turn them off, and that vehicle is definitely going to go somewhere.

And of course, the approach has failed utterly, in terms of delivering affordable and routine access to space.

The new private (and old governmental one, with the X-15) approach of building on suborbital capability is just the opposite, in which vehicles are tested incrementally, slowly expanding the envelope of performance, like an aircraft development. In this approach, the stress is on low cost from the beginning. As Jeff Greason, president and founder of XCOR Aerospace, has said, it’s easier to figure out how to do something reliably and affordably and then get more performance out of it, than to focus on the ultimate performance first, and then try to reduce its costs and increase its reliability.

Of course the first suborbital vehicles will not simply scale up to go to orbit. But that’s not the point. A vehicle that goes to a hundred kilometers altitude at three times the speed of sound can be scaled up to one that goes to two hundred kilometers at four times the speed of sound. And because it does it so affordably, it provides a cost-effective test bed for new thermal protection techniques, or propulsion or life support systems, to start to extend the performance envelope further with new vehicle designs, still maintaining low costs per flight. And if there are multiple companies building such vehicles, they’ll be able to learn from each other’s mistakes as well. Mach 5 can become Mach 7, Mach 7 can become Mach 12, Mach 12 can eventually become Mach 25 and orbit, as experience is gained, and designs evolve.

This is a radically different approach for space flight, but it’s one with which aviation enthusiasts will be familiar—it’s how aircraft technology advanced rapidly between the first two world wars. And moreover, it’s an approach that we might have been taken much earlier in space, had we not been diverted to the wrong track—away from incremental development based on flight testing—by the post-Sputnik panic. That panic, which culminated in the sixties with the race to the Moon, demanded "space spectaculars" to match the Soviet ones, even if that meant high-risk leaps into the technological unknown. This new (and at the same time old but untried) approach may not work, but it can’t be worse than what we’ve had for the past forty years.

The fundamental question, of course, is what will be the economic driver for it? Are there adequate private markets? Or will the government have to actively encourage it, and in more specific ways than motherhood statements in the Aldridge report?

In fact, the truth is exactly the opposite of Myth #5. Technology isn’t an enabler of public space travel. Public space travel, and new markets, becomes a financial enabler of the development of the new vehicles necessary to satisfy the demand. If one accepts the premise that low costs come from high flight rates, then absent any other large markets, large-scale public participation in space flight is essential to lowering the costs of access. Fortunately, we know, from numerous public opinion polls, that at least half the US population (and a higher percentage in Japan) wants to visit space. The only issue is the price, features and safety levels necessary to satisfy the various market segments. The best thing about this market is that, unlike other postulated space market drivers, such as solar power satellites, or space manufacturing, or Helium 3 mining from the moon, it requires no technological developments other than the space transports themselves—the payloads are in fact already built, with a simple interface to the vehicle (seat and keister), and they are self loading. And many more are being manufactured every day, by low-skilled labor, regardless of whether they ever make it into space.

Of course, we obviously have a chicken-and-egg problem here. Low costs are necessary to satisfy demand, and new vehicles are necessary to provide the low costs, and demand from a promising but still-unproven market is necessary to raise the funds to build the new vehicles. Fortunately, a small hummingbird egg is slowly starting to hatch, with the success of the X-Prize and the increasing interest from the public and investors. It may provide the beginning of a virtuous cycle of development that eventually results in hummingbirds, then sparrows, then chickens and more chicken eggs, and eventually eagles.

Incremental development undercuts the myth. In fact, there have been many deposits already taken for suborbital flights, and with the success of SpaceShipOne, it’s becoming clear that they can be offered at a price acceptable to the market. Yes, people will pay more for orbit, but that doesn’t mean they’ll pay nothing for suborbit. Clearly, based on hints dropped by Paul Allen and Burt Rutan about a SpaceShipTwo, with larger passenger capacity, and Jeff Bezos’ (founder of investment into his own suborbital venture, some non-naïve people believe that there’s a business case to be made. The Futron Corporation has performed market research indicating that there’s a market for at least several thousand tickets, even at a price level of a hundred thousand dollars.

In addition, as the vehicles become more capable, they may open up new markets for rapid delivery, of either people or cargo, from one continent to another. They may also find applications for potential military and remote sensing missions, and high-altitude or weightless research (the latter of which is satisfied now, at very high costs, with expendable sounding rockets). Glenn Law and Jared Martin of the Aerospace Corporation, funded by the Department of Commerce, published a report on this topic in October of 2002 (three months before the loss of Columbia), available at

What are the implications? The CEV is not planned to fly with crew until well into the next decade. Even ignoring its undoubted billions in development costs and the fixed costs of support for it, it will cost at least a hundred million dollars just for the expendable launcher to deliver it to orbit, with its tiny crew complement. This is ignoring the additional costs of the perennially misunderstood issue of “human rating” the launcher (yet another myth, beyond the scope of this essay).

If progress can occur sufficiently rapidly (and based on how quickly progress has initially been made, by Scaled Composites, Armadillo Aerospace, XCOR and others, that could be rapidly indeed in the context of the normal pace of government space programs), it’s quite possible that by the time, a decade from now, NASA’s single-point-failure CEV is ready to fly with its first crew of NASA astronauts, it may be superfluous, superceded by multiple vehicles capable of delivering humans to and from orbit for a tiny fraction of its costs.

Those same vehicles could deliver exploration components to space “dry” (that is, without propellants), and the propellants could go up on yet other launches to be delivered to depots in earth orbit (perhaps associated with orbital tourist hotels, like the ones that Budget Suites owner Bob Bigelow is spending his own hundreds of millions to develop). Ultimately, it’s conceivable that (at least facetiously), should the government ignore this new industry, and continue along the stasist path laid out by the president in January, the first NASA astronauts back to the moon since 1972, sometime around 2020, could be greeted by the concierge at the Luna Hilton.

Of course, in reality, if the private sector is making rapid progress, with or without government nurturing, it will become clear long before CEV flies to orbit with a crew, let alone to the moon, that the policy has to be adjusted to the new reality. The real issues are:

  1. Is the thesis correct and;

  2. What should the nation’s space policy be if it is?

The thesis can’t be defended any further here, short of expanding this essay into a book, but here’s a suggestion of what the shape of an appropriate policy might be, assuming that a convincing case has been made.

Without getting to deeply into specific details (wherein is the lair of the devil), here are some potential areas in which different (in some cases radically different) government policies could help:

While the administration should be applauded for coming up with new policy, the secretive means by which it occurred didn’t provide ample input from more diverse viewpoints. The outreach activities of the Aldridge Commission mitigated this to some degree, with its public hearings, but the commission was still working within the confines of the fait accompli of the president’s initial proposal. It’s been over three decades since humans last walked on the moon, and even under the president’s proposal, it’s likely to be at least another one before they do again. It’s not unreasonable to step back, now that we’ve established that we’re going to move out into the cosmos, and accept a little more delay as we try to determine how best to implement a policy to make that happen. Not just to make it happen, but to make it happen for all of us, not just NASA astronauts, even if it doesn’t happen exactly in the way that government five- and ten-year planners might desire.

The president came out with a new policy in January because he recognized that our nation is at a crossroads. The loss of the Columbia last year made very clear that we could not continue into the future on the policy inertia of the failed past. But we still don’t seem to recognize the road markers. It’s not a choice between old government space programs—Shuttle and station—and new ones with the “right technical approach.” The choice is more fundamental.

We can remain constrained by the old familiar myths, or we can cast them off and take a fresh look at space policy unblinkered by them. We have an opportunity now to return to the dynamist path that we might have taken, had we not been diverted by the need to defeat communism so long ago.

We can build a new space age on the ashes of the old, a space age whose characteristics aren’t central planning and command, but organic growth, not one for a few public employees but one rather for the many private citizens. It can be one not just about science and national security (and federal pork), but one more in keeping with the growing wealth of the planet, one about education and entertainment and a literal broadening of horizons. Perhaps most importantly for Americans, it will finally be a space age comprised of the more traditional values that opened up other frontiers in the past—not collectivism, but individualism, free enterprise, and freedom in general.

Perhaps, decades from now, our descendants (and even we ourselves if current trends in life extension continue) will look back on the earth from some distant planet, and wonder how, for so many decades, we fooled ourselves that it could have been done any other way.