NASA abandons Hubble:
an incarnation of "Lifeboats on the Titanic" thinking

by Roger Bourke White, Jr., copyright February 2004


I have been reading in the news that NASA has decided to abandon the Hubble Telescope. [After the Columbia Shuttle Disaster in 2003] It is doing so because under the new safety guidelines for the space shuttles, the shuttles can't make the flight to the Hubble safely.

According to the new safety guidelines, the shuttles can only fly to places where they can be docked so they can be inspected and some repair work done on them, if necessary. There's only one place in space where the shuttles can dock, so the only place in space the shuttles can fly to is the International Space Station.

What a shame, and what a perfect example of the "Lifeboats on the Titanic" phenomenon.

The Lifeboats on the Titanic

When the Titanic crashed into an iceberg and sank in the Atlantic in 1912, a scandal followed because so many lives were lost on what was supposed to be a very safe journey. The hue and cry went up to do something to make the transatlantic crossings safer.

One of the measures invoked to improve post-Titanic safety was instinctively obvious: Many people died in the sinking because there weren't enough seats in the lifeboats for all the people on board. So, one of the recommendations was that all passenger liners be outfitted with enough lifeboats so that all the people on board can have a seat.

What a noble and intuitively right thing to decide! But in practice, how expensive and ineffectual! Why do I say this? Well, first off, if it was that practical as a safety measure, why wasn't it being done before the Titanic sailed?

How do ships crash, and what are lifeboats used for?

The answer to why it wasn't being done before the Titanic crash is found in how lifeboats are used in a typical shipwreck.

The average shipwreck:

Why are these conditions important? Because they all affect how many lifeboats can be used and how they are used.

First, a crash happens because something has gone wrong, and most times what has gone wrong is weather related. This means most wrecks happen in rotten, stormy weather, and there's going to be wind and waves bashing into the ship during and after the wreck. These crashing waves and high winds are going to be even more hazardous to any lifeboats lowered than they were to the mother ship, so unless the ship is in imminent danger of sinking, the captain and crew will wait for the weather to improve before launching lifeboats.

Second, if a ship crashes near a coast, the lifeboats will be used to ferry people from the wrecked ship to the coast. This means that everyone doesn't have to be on the lifeboats at the same time.

Third, if the ship tips over when it crashes (lists) -- and this happens more often than not -- then half the lifeboats (those on the high side) will not be launchable at all. This means that in most wrecks that have to be abandoned quickly (those that can't wait for the storm to end), having a lifeboat per person means that only half the passengers and crew will have a seat.

Fourth, as mentioned in number one, if lifeboats are going to be launched quickly, they will be buffeted by the same wind and waves that the mother ship was buffeted by, and they will have to dodge the same rocks that the mother ship has crashed on. If the lifeboats can't survive these conditions any better than the mother ship could, there is no point in launching them.

(Update: Here is the description of a more typical memorable cruise liner shipwreck, the Costa Concordia in 2012. Note how the lifeboats are used.)

The Titanic sinking was a freakish accident in many ways:

It was only under these freakish conditions that not having one lifeboat seat per person made any difference at all, yet in spite of its one-in-a-million nature, the Titanic accident set the standard for lifeboat capacity for the next hundred years, and we live with this extra lifeboat expense to this day.

This producing of a safety policy based on its intuitive feel rather than its formal cost-benefit relation/ratio is what I have labeled the "Lifeboats on the Titanic" phenomenon.

Flashforward to the Columbia

In 2003, the Shuttle Columbia broke up into little pieces over Texas, and NASA's space program was scandalized for a second time. (The first was when the Challenger blew up shortly after launching in 1986.) For a second time, the shuttle program was shut down and an investigation ensued.

The investigating committee decided that the cause of the breakup was a hole punched in a wing by insulating foam crashing into the wing on launch.

Having decided that was the cause, NASA changed the safety policies for the shuttles so they could fix that particular problem if it happened in the future. The cost of doing so, as mentioned at the beginning of this essay, was that the shuttles would no longer fly anywhere but to dockable destinations, i.e., the Space Station.

As with the Titanic, the question that has to be asked is: Is the benefit of the improved safety worth the cost of what is being sacrificed to gain it? In the case of the Titanic, the answer after nearly a hundred years of experimenting is no -- no other passenger liner has ever shipwrecked like the Titanic did, so no people have ever been saved by the money spent on extra lifeboats.

In the case of the shuttles, the question has to be: What is the likelihood that some future catastrophic accident will be prevented by something the crew can find and fix up in space? If it's one-in-a-million, then why are we sacrificing the Hubble? This is especially poignant because the same article says the director of NASA wants to retire the entire shuttle fleet in a few years. If that's so, then we are talking about ... what ... a hundred or so flights? A hundred is one ten thousandth of a million, so what are the chances of our one-in-a-million accident happening again?

This decision to sacrifice the Hubble is another example of lifeboats on the Titanic thinking: An example of a safety action being taken on the basis of intuition and emotion, not rigorous cost-benefit analysis.

Such decisions, while comfortable at the time they are made, are always very expensive for humanity. We should, instead, recognize that we live in a risky world, and recognize that it is not cost-effective to spend on trying to eliminate one-in-a-million accident scenarios, unless we expect that scenario to be repeated at least a thousand times within a reasonable time frame. We would then be spending to prevent something that had a thousandth of a chance of being seen again -- still a vanishingly small risk. (Epilog: Cooler heads did prevail and NASA did make one exception and visited the Hubble a final time in 2009.)

Update: This Feb 2012 article in Reason magazine, How Much Is an Astronaut’s Life Worth? by Robert Zubrin, describes the alternative to the emotional approach. It's an analytic, cost/benefit approach to deciding when to risk a Mars mission. The result of using this style of analysis is a much clearer, and much quicker, decision on when to go ahead.

Update: The anniversary of the Titanic has produced this 12 Apr 12 WSJ editorial, The Real Reason for the Tragedy of the Titanic by Chris Berg, which summarizes the various reasons that have been given over the years for why there weren't more. "The disaster is often seen as a tale of hubris, social stratification and capitalist excess. The truth is considerably more sobering." It points out what I have said in this article, that the expected use for the lifeboats was ferrying passengers to nearby rescuing boats or shores, and this was based on the precedent of preceding accidents.


--The End--