Wednesday, January 6, 2010

Security and mathematics.

In light of recent attempted plane bombing...
I think there is a huge number of unaccounted deaths from September 11 2001 attacks, in range of thousands, and I'm not talking of Iraqis or US soldiers. I'm speaking of US civilians in the US.

Let's look at US traffic fatalities (about 42,000 per year)
Note the rise in fatalities after September 11 2001. The traffic fatalities before 2001 were going up and down, and since 2001, only gone down in 2007 due to rising oil prices, and gone down sharply in 2008 due to the recession.

I'm pretty sure that plane security did increase the air travel time, as well as fear of flying. Consequently, some people had opted to get around interstate by car, especially for shorter distance flights where most of time is spent boarding the plane. Cars, as you can see, are significantly less safe than airlines even in the worst year (2001) counting the on-ground causalities.

I cannot know what percentage of people opt to drive rather than fly. That would require a poll. What I know is that even very conservative, small increase in car usage (1%), when multiplied by 42,000 and by 6 years, gives fairly substantial number (2,500), approaching the direct causalities (2,974 according to Wikipedia). Looking at the traffic accidents statistics, it could be even worse; thousand extra deaths per year.
That is seriously depressing. Aftermath of subway attacks in Britain may be even more depressing.
If you agree with the argument, and care about it, try make the difference.

Before giving up liberties and conveniences to avert some risks, it is absolutely essential to do a proper cost benefit analysis; failure to do so is quite equivalent to cooperation with terrorists, to direct participation in mass murder, for the suboptimal action easily results in huge number of additional deaths. Sometimes it may even be the case that very minor response is best (e.g. reinforced cockpit doors as response to 9/11/2001 could well be sufficient to prevent it from ever happening again, not because its absolutely impossible to cut through reinforced door somehow but because its secure enough to point that terrorists would rather do something else instead).
The problem is, how can democratic society do the cost-benefit analysis? A company does the cost benefit analysis for itself; for media companies, for military contractors, for pretty much all businesses involved in security, for members of the government, for many security experts, there are huge benefits from overreaction to the risks. Media is working to amplify our fears out of proportion; you will never see a major media channel comparing risks - unless it is e.g. promoting 'war on drugs' and wants to convince you that war on drugs deserves as much attention as war on terror.
On other hand, entire society bears the costs and extra risks resulting from the overreaction, not just the individuals getting the benefits. Consequently, response to terrorism is always overreaction, always biased for those benefiting from overreaction, against those harmed by overreaction.

Consider a bee sting. Normally, it takes hundreds, or thousands killer bees to kill healthy adult through the sheer amount of venom. But some people have highly overzealous immune system, sensitized to the bee sting, going out of control and destroying their own cells. Such people often die from a single sting - condition known as anaphylactic shock - people often get killed, in effect, by their own immune systems. Immune system is not intelligent - but we are. We should not let this happen to our culture.

We should not turn this failed plane bombing attempt into a successful act of terrorism, which it could easily become if boarding checks for internal travel are to take more time, for short distance flights, where driving, a much more dangerous alternative option, would become quicker and preferable leading to extra deaths in car accidents which would have been mostly prevented if people would travel by plane instead (airlines are a lot safer per km traveled than cars, even for short distance flights).

What is especially illuminating is those full body x-ray scans. I'm fairly sure that there had been no evaluation whatsoever of the number of extra cancer deaths that may be caused by x-ray scans versus deaths from terrorism that could be, arguably, prevented with those scans, which are unable to detect explosives within the body or detect plastic explosives reliably. All you can see in news about those machines is talk that dose is "very small", without actually naming the dose, and its hard to find dose estimate from reputable source (indeed, from any source other than equipment manufacturer). I bet if you ask a technician, he would be quite unable to name the dose as well (total incompetence is amazingly common).
It's as if prices did increase, and the increase was not described in percent but simply as "small". You can't straightforwardly measure x-ray dose with Geiger counter, not for low energy x-rays, due to low rate of detection for xray photons. (Though, I don't think extra x-ray dose is important problem in comparison with more car driving rather than flying because of people not wanting to be seen nude, deaths from money being spent on this rather than on healthcare, and so on and so forth). Other take on the issue.

No comments:

Post a Comment