Here is an article warning about the next big earthquake in the San Francisco Bay area. When I lived in Oakland I got used to seeing articles like these two or three times a year, particularly right after a small but noticeable earthquake.

Two things in the article catch my interest. One is the last word.

One of my pet joys in reading is to encounter a word used in a way that is completely consistent with the ordinary common meaning but at the same time is in accord with its older, core meaning. To isolate something, in its original sense, means to make it an island. (Its near-synonym, insulate, has the same parentage: insula is island in Latin; isola is island in Italian.) Here, I imagine those building being literally isolated, as the sinking ground leaves them as islands in the Bay.

The other is this:

This sort of thing has been a peeve of mine ever since my days working at Underwriters' Report, a now-defunct trade magazine for the insurance agency.

What bothers me is that damages are measured in dollars. I can't fault the sentence for what it says. It's true. The one sum of damages does far exceed the other. But whenever statements like these end up in a newspaper article, the implication to the reader is that the one scale of destruction also far exceeds the other, which doesn't necessarily follow.

Property values in the San Francisco area are among the highest in the country; property values in New Orleans, even before the hurricane, were not. If the identical office buildings were built in downtown San Francisco and in downtown New Orleans, and then one was utterly destroyed by an earthquake and the other was utterly destroyed by a hurricane, is it really fair to say the earthquake was more destructive, simply because office space costs so much more in San Francisco?

When I see damage forecasts measured in dollars, I have to ask how much of the difference is simply a measurement of property values. Alas, my googling attempts have failed to turn up statistics on property values in New Orleans before the hurricane. If you can find them for me, please share. Until then, I'm settling for the current consumer price index, which I think will serve as a rough proxy. New Orleans isn't listed anymore, so I substituted Houston, which again I'm hoping is reasonably close.

I don't remember which baseline I chose, but since I'm just comparing the one number to the other, it makes no difference so long as I used the same baseline for each, which I did. The numbers were 216.04 for San Francisco-Oakland-San Jose and 183.84 for Houston-Galveston-Brazoria, for a ratio of 1.18 to 1. That's nearly identical to the 1.17-to-1 ratio of the damage forecasts.

So by this measure, the numbers are really saying that the expected property damage from the earthquake will be about the same as Hurricane Katrina. To be more exact, you'll have to adjust for my substitutions: Houston vs New Orleans, consumer price index vs actual property prices. While you're at it, you should probably consider that the dollar figures in the article are probably today's dollars for the earthquake and 2005 dollars for the hurricane.

But the point remains. The article strongly suggests that damage from the anticipated earthquake will be even greater than damage from Hurricane Katrina, but unless your definition of "damage" is claims in dollars that the insurance companies will have to pay, the numbers really don't show that.

11:34:02 PM  [permalink]  comment []