YOU ARE HERE: LAT HomeCollections

Statistics and Terrorism

September 18, 2004

Re "Terror Threat May Be Mostly a Big Bluff," Commentary, Sept. 13: Bart Kosko's comparison of the terrorist threat to the Loch Ness monster and Bigfoot was incredibly insulting to the thousands of families who lost loved ones on Sept. 11, myself included.

I was in fact shocked The Times would publish such a shallow analysis by someone with no credentials in the field of international security.

Kosko's absurdly academic approach fails to address the crux of the issue: We fear terrorists because a single cell can cause unimaginable destruction.

As witnessed most recently in Russia, the enemy is a death cult acting outside the bounds of human decency.

Worse, the terrorists are perilously close to obtaining the technology of mass murder (see the anthrax attack, and Abdul Qadeer Khan's nuclear bazaar in Pakistan).

A single biological, chemical or nuclear strike would render Kosko's phony statistical reasoning obsolete.

It is precisely the asymmetry of the threat that the country fears -- and it is nothing like the exaggerated fear of a shark bite. Last I checked, no shark ever accounted for 2,900 deaths with a single bite.

Jonathan Gardner

Beverly Hills


I read with great interest your commentary by USC professor Kosko, not only because this is a timely topic but also because I was one of his students in the USC engineering master's program.

Kosko is a brilliant statistician, but the one scenario he does not address is the one in which the cost function has a prohibitively high penalty for a Type I error.

In layman's terms, what if a terrorist acquires a so-called nuclear backpack bomb or a dirty bomb and detonates it in one of our cities? The loss of life would not only be catastrophic but the event could bring down our entire economy if it were done in New York or Los Angeles.

If the probability of this event were even a little higher than remote, then the probability times the cost would absolutely dominate the expected value of a terrorist attack. In statistical terms (which I learned from Kosko), E(X) would be high. In layman's terms, we are living under a very serious threat of attack.

Joseph Gold



Kosko made an interesting case about the threat of terrorism being overestimated. But what would he have the government do? If another terror strike hit, would he simply put a tally in a column and think of more statistical probabilities?

As President Bush and others (but not Kerry's team, because acknowledging reality would force it to say Bush has done well) have said, America's intelligence and security have to be good 100% of the time. A terrorist has to get through only once. And the consequences, as we saw on Sept. 11, can be severe. The problem for Kosko is that human lives are not statistics. Fortunately or not, I don't live in a probabilistic world. I live in a real, physical world, where a bad guy can inflict real (not probable) harm.

Do probabilities have a place in our evaluations? Sure. But to just write off the war on terror as a statistical artifact without a shred of evidence does us a real disservice.

Bal Simon

Bellevue, Wash.


Kosko's expose of the myth of terrorism as a bluff is something I have thought all along: There is no threat of terror nor has there ever been one. It is a ploy by Bush and Co. to keep the atmosphere of fear going.

The administration keeps Americans in a perpetual state of angst, much like an abusive parent, creating an even greater dependence on the abuser. This dependency became clear to me in a photo exhibit some years back at Barnsdall Park, showing pictures of families who had abused their children. Photos were taken at precisely the moment the abused child returned home from foster care, and they always showed the child running into the arms of the abusive parent, or in psychological terms, identification with the aggressor. There is no war on terrorism. Just aggressive acts of violence by this nation against a nation that never bothered us.

Naomi Stephan



Thank you for Kosko's piece. Amid all the emotionalism and politicking around this issue, it's nice to read a perspective that takes a more distanced view.

I just wish it might have been titled "Car Accidents Kill 40 Times as Many Americans as Terrorists." In fact, I wish The Times would make it a headline on the front page.

We are much more in denial about car accidents than we ever will be regarding terrorists.

Tom McGee


Los Angeles Times Articles