The Space Reviewin association with SpaceNews

Previous comparisons between the privately-developed Falcon 9 and India’s government-developed GSLV fail to take into account the record of success built up by India’s PSLV. (credit: ISRO)

Picking up the torch vs. dropping the ball

Bookmark and Share

Last week Dr. Sam Dinkin wrote an article for The Space Review that includes inaccurate numbers, faulty logic, and illogical and misleading analogies in an effort to argue that private wealth will inevitably lead to space development. Titled “Individuals pick up the space development torch,” (The Space Review, June 21, 2010), the article makes an argument that many people are sympathetic to, and therefore may lead them to miss its substantial flaws.

The article contains several themes:

  • Private companies are more efficient at developing spaceflight than government entities
  • Private companies are better at developing spaceflight than government entities
  • The increase in the number of rich people around the world makes private space development virtually inevitable

Although there is evidence to support some of these claims, the evidence provided in the article is lacking. Somewhat oddly, the article attempts to support them by comparing Elon Musk’s personal investment in SpaceX with several other expenditures that are not equivalent: the reputed development cost of the Redstone ballistic missile, the reputed launch cost of an Ares 1 rocket, and the annual budget of the Indian Space Research Organization.

Private efficiencies

It is a core belief of American capitalism, particularly since the Reagan era, that private industry is more efficient than government at producing everything from hammers to airplanes. There is in fact evidence to support this conclusion and a number of studies that have evaluated it. Usually those studies also seek to explore the causes of this disparity, sometimes concluding that lower government efficiency is a result of the compromises inherent in democratic government (for example, the necessity of seeking broad support for an expenditure), and sometimes concluding that the government trades lower efficiency for some other desired factor (such as classifying procurement in the interests of national security).

It is a core belief of American capitalism, particularly since the Reagan era, that private industry is more efficient than government at producing everything from hammers to airplanes.

Despite a significant body of literature on this subject, the article relies upon a few poorly applied analogies and inaccurate comparisons to try to make this point. For example, the article starts with a curious comparison of computers with rocketry, and in both subjects the article makes simplistic comparisons to try and demonstrate that private entities have been able to achieve important technological developments with less money. In the area of computing, the article states that the US Army developed the ENIAC in 1946, which cost $6 million in inflation-adjusted dollars, whereas the first Apple I was privately-developed in the mid-1970s for $5,000. The Apple I led to the Apple II, which “democratize[d] computing” according to the article. In the area of rocketry, the article compared the Redstone, developed primarily “by General Electric and the Army” for $700 million in inflation-adjusted dollars, with SpaceX’s Falcon 1/9, developed “with $130 million investment.”

Although the article notes that “the hard cost of development of rockets has not fallen” by the same factor as computers, these examples represent a theme that permeates the article, which is that private entities spend less money to achieve more than government entities do. This may be true, but the examples in the article never compare equivalent activities in order to demonstrate that this is the case.

For starters, both the ENIAC and the Redstone were developed when their respective technologies were new and immature—of course they will have higher development costs in inflation-adjusted dollars than similar objects developed decades later in mature industries. So while such comparisons can teach us lessons, they don’t really demonstrate the superiority of government vs. private development.

Similarly, in both computing and rocketry, the later private developments benefited from decades of government investment. Could Steve Wozniak have built that device in his garage if the government had not invested in computers and transistorized technology during the Cold War? Could the Apple I have existed without the pioneering work on the ENIAC? Similarly, SpaceX did not simply pop into existence and develop a cheap, effective rocket. The company benefits from decades of government-funded rocket (and computer) development, not to mention employees who were trained on previous government-funded aerospace programs.

Of course, the article was also trying to demonstrate another point: the falling development price of these technologies. But unfortunately the argument was undercut by flawed math: the statement that Falcons 1 and 9 only required $130 million “investment.” Elon Musk stated soon after the Falcon 9 launch success that the development cost was between $350 and $400 million, much of which came from NASA. What is the point of comparing the development cost (supposedly $700 million) of the Redstone with a single investor’s stake—and not the similar development cost—of the Falcon family? This is not an apples-to-apples comparison. At best, it is a factual error, but it could also be seen as trying to make a point by cooking the numbers and selecting the lower investor’s figure, rather than the higher development cost.

In both computing and rocketry, the later private developments benefited from decades of government investment.

Lest you consider this a minor error and not a fundamental flaw, it is worth noting that later in the article there is a similar apples-to-bathtubs comparison of the theoretical cost of a single Ares 1 rocket launch (allegedly $1 billion, although it is not clear where this figure originates), with the “development of a whole new SpaceX-style firm and rocket,” which is here stated as $120 million (vs. $130 million as stated earlier in the article). Why is Elon Musk’s investment in SpaceX listed as $130 million at one point, and the SpaceX development of the Falcon 1/9 listed as $120 million at another point? And why is the investment price compared to the development cost of the Redstone, and later the development cost compared to the single launch cost of the Ares 1? The reader can be forgiven for confusion, because the article itself is confused.

There are several other notable errors in the article’s second paragraph alone: Redstone development actually started in the early 1950s, not 1946, and it was produced by Chrysler, not General Electric.

In making these comparisons between computers and rocketry, the article assumes that there are only a few relevant variables: 1) inflation, and 2) the nature of the entity spending the money, i.e. “government” vs. “private.” But this is where that $130 million investment vs. $350–400 million development cost becomes far more important. The reality is that the additional money SpaceX used in Falcon 1/9 development came from government contracts. Thus, a comparison between a “government”-funded Redstone and a “privately”-funded Falcon family falls apart. Maybe SpaceX could have developed the Falcon 1/9 without several hundred million dollars of government funding (not to mention the promise of more). But the reality is that the company received the money and used it, and the company is competing to get more.

Later, the article states that “government-developed super computers continue in the age of privately-developed microcomputers,” but “have become increasingly irrelevant to the economy compared to privately-developed computers.” Although a true statement, it misses the point, which is that the government-developed super computers are not intended to serve the economy, they are intended to serve national security, or other government needs, such as weather forecasting.

“Government dropped the torch”

The section headlined “Government dropped the torch” contained more flawed logic. This section seems to claim that the relevant variables are: 1) “government” vs. “private,” and 2) launch failure vs. success. But, as noted, SpaceX has substantial government funding. SpaceX also benefits from substantial government R&D investment in rocketry, spaceflight and education over many decades. In addition, SpaceX also benefits from not having to pay full costs or amortized costs for things such as the launch ranges that it flies from. If one is going to try to argue that “government = inefficient and therefore bad” vs. “private = efficient and therefore good,” then a little more nuance, complexity, and honesty is called for.

Once again, the article makes some rather disingenuous claims. For example, in this section, two failures of South Korea’s first launch vehicle are compared to SpaceX’s successful launch of Falcon 9. “SpaceX had some launches that were not deemed as successful as their most recent one,” the article states (emphasis added). Put another way, they did not reach orbit, just like South Korea’s KSLV launches. Later on, SpaceX is compared to the Ares 5 paper rocket, which is said to be “in the process of failing before it even reaches the launch pad.” So, according to this article, a “failure” is something that happens when a government-developed rocket falls into the ocean, or when a government program is canceled by a new political administration, but apparently the word “failure” cannot be used when a “privately-developed” rocket falls into the ocean. The reader can be forgiven for thinking that this language might be a bit biased.

When the probability of success is so low and the chance of failure so high, it is a mistake to draw major conclusions.

There are also two problems of logic in this section of the article. The first is trying to draw assumptions from such incredibly small sample sizes, as the article seeks to do by using examples of recent rocket launches to prove that private industry is more successful than government at developing rockets. When N=1 or 2, how can we make effective comparisons? Because South Korea had two launch failures with its KSLV rocket and SpaceX had one success with its Falcon 9 can we then immediately jump to the conclusion that this proves that a private company is “better” than a government at developing a rocket? If the Falcon 9 had failed on its maiden launch, would this therefore be proof that private industry cannot develop rockets?

By Elon Musk’s own admission, Falcon 9 had at least a 20% chance of failing on its maiden flight. It succeeded. But what were the chances of failure for the other rockets mentioned in the article? If the South Korean rocket also had a 20% chance of failure, it is entirely possible that it was simply unlucky during its launches. When the probability of success is so low and the chance of failure so high, it is a mistake to draw major conclusions.

But the article has also badly distorted the record in order to try and make this point. For example, the article states: “India, South Korea, and NASA have expensive rocket developments that are not as successful as SpaceX’s.”

Let’s take India, for starters. India has been launching rockets since 1979. The country’s first Satellite Launch Vehicle had one failure, one partial success, and two complete successes. It formed the basis for the Augmented Satellite Launch Vehicle (ASLV), which started flying in 1987 and suffered two failures, a partial success, and a complete success. Because of a lack of funding, India discontinued the ASLV in favor of the Polar Satellite Launch Vehicle (PSLV), which started launching in 1993. The PSLV had three development launches, one of which failed, and thirteen operational launches, one of which was only a partial success (reaching the wrong orbit). India has also developed the Geosynchronous Satellite Launch Vehicle, which has had six launches to date, with two successes, two partial successes (wrong orbits), and two failures.

Now compare that to SpaceX’s record: six launches, three failures to reach orbit.

The only way that statement (“India… [has] expensive rocket developments that are not as successful as SpaceX’s”) is true is if you compare only India’s GSLV record and only SpaceX’s Falcon 9 record. If you include the Falcon 1 in the comparison, then the two are essentially equivalent—as long as you decide that the GSLV, which has a cryogenic stage and reached geosynchronous orbit, is equivalent to SpaceX’s less capable rockets. But this again highlights the problem of small number comparisons: if the inaugural Falcon 9 launch had blown up—i.e. a single failure—there is no way that this comparison would be true.

Looking at it another way, however, India has launched sixteen PSLV rockets with fourteen complete successes. In fact, India has successfully launched twelve PSLV rockets in a row over a period of eleven years. If one is trying to argue that a government rocket development is inferior to a private rocket development, it helps to ignore the PSLV, as the article chose to do.

But by this part of the article, it is hard to believe that these errors and logical fallacies are either minor or unintentional. The article once again trudges up SpaceX’s “startup capital” (i.e. the $120 [or $130] million figure stated elsewhere) and compares it to the $1.2 billion annual budget of the Indian Space Research Organization, although the point of such a meaningless comparison is unclear. ISRO spends its budget on a lot of things, including multiple rocket launches, scientific and applications satellites, ground stations, research centers, and technology development. Why should the total cost of all these endeavors be compared to the investment by one American entrepreneur in one company?

Economic and cultural determinism

On several occasions in the past, the author has advanced theories that can essentially be described as economic determinism—the view that current economic trends, extrapolated far into the future, will produce a predictable outcome. For example, in 2007, he wrote: “Per capita GDP will be about 1,000 times what it is today and the median family will be able buy a ticket off of Earth at today’s prices.” (see “The ultimate solution to global warming: emigration”, The Space Review, July 16, 2007) In October 2009 he wrote that based upon economic factors alone, interstellar travel and even the colonization of the galaxy were inevitable (see “Clinical immortality and space settlement”, The Space Review, October 26, 2009). In the most recent article, the author argues that growing global income, growing percentage of income that is capitalized as wealth, and shrinking cost of rocket development “are leading to a growth in the number of people with the potential to build an orbital rocket of over 10 percent per year.”

There is no reason to believe that private spaceflight investment or development, or even private spaceflight itself, is inevitable, because private interest in spaceflight is a cultural phenomena just as much, or perhaps more, than an economic phenomena.

Now it is a staple of post-Enlightenment thinking, and particularly American optimism, that the human condition continues to improve. This is easily demonstrated by readily available data. We eat better (indeed, eat too much), live longer, earn more, have more disposable income, more leisure time, etc., when compared to previous generations. Despite short-term setbacks, and concerns about developments like the accumulation of substantial government and personal debt, or possible environmental degradation, it seems a reasonable bet that these overall trends will continue. They will not necessarily continue at the same pace in the past, but there are more reasons to believe that an American citizen of 2050 will be better off than a citizen of today than they will be worse off. Very few people believe that social and economic collapse is in our future.

But there is no reason to believe that these trends make private spaceflight investment or development, or even private spaceflight itself, inevitable, because private interest in spaceflight is a cultural phenomena just as much, or perhaps more, than an economic phenomena.

Today, compared to 2,000 years ago, there are many more individuals who could build pyramids… and yet, there are not many millionaires or billionaires constructing pyramids. Of course, the response to this is that there are many rich people building the equivalents of pyramids—Trump Tower(s), massive yachts, personalized jumbo jets. But pyramids themselves are a culturally outdated concept, and therefore despite all the rich people, the world is not teeming with pyramids.

The current crop of private space investors was influenced by the space race and its cultural artifacts, such as Star Trek. They have sufficient extra income to spend it on these interests. But there is no reason to believe that, in the future, rich people will want to spend their money in the same way that some rich people, like Elon Musk, Robert Bigelow, or Jeff Bezos, are spending their money today. Twenty or thirty years from now there will be a new generation of rich people looking to invest their money, and they will have their own set of cultural influences that may lead them to invest that money in things other than spaceflight.

What will that be? What current cultural influences today are forming the dreams of tomorrow? Videogames and the simulated worlds they depict seem to be a likely cultural influence. Maybe some fifteen-year-old boy who is currently obsessed with playing Bioshock 2 will grow up to be a billionaire obsessed with building an undersea city, rather than a cheaper rocket. Maybe the future wealthy people participating in games like World of Warcraft will in the future seek to develop more realistic immersive environments, or genetically-engineered dragons. And maybe dreams of spaceflight, fueled by current wealth, will soon fade and be replaced by different dreams entirely. Certainly spaceflight no longer holds the imagination of the American public like it did decades ago.

The point—at the risk of belaboring it—is that current interests, and current economic trends, are not deterministic. Dr. Dinkin has drawn a line on an economics graph (i.e. society will continue to produce more and more rich people) and another line on a cultural map (i.e. people will continue to be interested in space travel) and assumed that these lines continue out into the future and continue to cross. Alas, the logic of this argument is also flawed.