A disaster of technology

26 years ago this week, the space shuttle Challenger disintegrated 73 seconds into its launch, leading to the death a couple of minutes later of its crew, which included a civilian, elementary school teacher Christa McAuliffe.  The disaster was replayed endlessly on television, becoming for my generation something akin to what the assassination of President Kennedy was to the previous one, i.e., everyone can remember where they were when they heard the news etc.  (I was in the office of the magazine where I worked, reading the acceptance letter I had opened thirty seconds earlier from the law school I would attend that fall, when a co-worker so rudely interrupted my reverie).

The event was investigated by a government commission, which, in the manner of such things, did its best to whitewash a series of administrative miscalculations which led to the accident.   Unfortunately for those administrators, one of the commission's members was the famous physicist Richard Feynman, who threatened to resign if he was not allowed to add an appendix to the official report.  This appendix makes fascinating reading for anyone interested in risk assessment and management, and in particular their relation to the politics of bureaucratic decision making.

The Challenger broke apart in flight because one of the o-rings which sealed the joints in the solid rocket boosters failed.  This failure was quite predictable, not merely in retrospect, but indeed in prospect, as the engineers who designed the boosters were aware.  Several of them had made concerted efforts to get NASA administrators to address the issue, but had been ignored.  The engineers estimated that the odds of a catastrophic launch failure were about 1 in 100, while NASA administrators insisted that the odds were more on the order of 1 in 100,000.  Feynman:

What are the causes and consequences of this lack of agreement? Since 1 part in 100,000 would imply that one could put a Shuttle up each day for 300 years expecting to lose only one, we could properly ask 'What is the cause of management's fantastic faith in the machinery?'

 NASA officials argue that the [risk of failure] is much lower. They point out that since the Shuttle is a manned vehicle . . . "the probability of mission success is necessarily very close to 1.0." It is not very clear what this phrase means. Does it mean it is close to 1 or that it ought to be close to 1?

In other words, the historical reasoning  process of the people running NASA went something like this:  Early unmanned rocket launches had very high failure rates.  With the maturing of the technology the failure rate declined quite a bit.  Once it declined enough, the risks were considered tolerable for manned flight ("tolerable" in the context of the cold war space race that is).  It was still understood to be ultra-hazardous, hence astronauts had to be extraordinarily brave people.

But after a few dozen successful manned launches over the course of a couple of decades, the administrators managed to talk themselves into the belief that space flight was "safe," in much the same sense that commercial air travel is safe.  This belief was absurd, but they had to maintain it if they were going to engage in publicity stunts, crucial to continued program funding, such as putting an elementary school teacher in space.  So they essentially reasoned backwards: it would be irresponsible to do what they were doing unless the odds of disaster were very low; they were doing this; they were responsible people; therefore it followed that "the probability of mission success is necessarily very close to 1.0."

Feynman's conclusion is particularly interesting, given the subject matter of this blog.  He argues that, given practical limitations, the overall risk of failure for Shuttle flights was going to have to be roughly on the order of one per cent (in fact it ended up being 1.48% over the history of the program).

Official management, on the other hand, claims to believe the probability of failure is a thousand times less. One reason for this may be an attempt to assure the government of NASA perfection and success in order to ensure the supply of funds. The other may be that they sincerely believed it to be true, demonstrating an almost incredible lack of communication between themselves and their working engineers.

 In any event this has had very unfortunate consequences, the most serious of which is to encourage ordinary citizens to fly in such a dangerous machine, as if it had attained the safety of an ordinary airliner. The astronauts, like test pilots, should know their risks, and we honor them for their courage. Who can doubt that McAuliffe was equally a person of great courage, who was closer to an awareness of the true risk than NASA management would have us believe?

Feynman, in other words, is making an argument for transparency.  Space flight is ultra-hazardous, which means that, to make undertaking it morally acceptable, those who do so should know the risks, rather than being misled by administrators who on one psychological level know they are lying to their audience, while on another level they maintain a sincere belief in their own rhetoric (Feynman's argument is psychologically sophisticated, acknowledging as it does that human beings are perfectly capable of maintaining contradictory beliefs simultaneously).

Of course in a sense analogizing risk management in the context of space flight to risk management in the context of trying to become a lawyer is, it might be said, a somewhat strained analogy.  Astronauts are perhaps the ultimate examples of risk-seekers, while traditionally law students have been classic risk-avoiders (I'll go to law school instead of trying to write the Great American Novel etc.).    The biggest irony of our own particular technological disaster is that law school -- the purportedly "safe" bail-out option for many a liberal arts major -- has been transformed gradually over the last couple of decades into, in economic and social terms, an ultra-hazardous activity.

Meanwhile, law school administrators, when surveying their bloated kingdoms, continue to cite overall "employment" rates as evidence for the proposition that the probability of career success is "necessarily" very close to 1.0. In this context, Feynman's rhetorical question -- does this mean it is close to 1.0 or that it ought to be close to 1.0? -- is becoming increasingly easy to answer.