Still responding to the last disaster
April 6th, 2007Posted by: admin
Eric Berger, the Houston Chron’s SciGuy has a Q&A up with David Paulison, the current FEMA chief. There are some interesting things in there:
Q. At some point do you advise someone in the federal government that programs like federal flood insurance should be revisited?
A. We need to re-look at the whole flood insurance program itself. How we provide flood insurance, what we’re going to charge for it, what requirement we’re going to have to get flood insurance. I don’t have all the answers for that right now, I can tell you that.
From what I’ve seen, the main problem is that NFIP is not allowed to be a true insurance market because political interference from Congress will not allow NFIP to charge actuarially-sound rates. Maybe Mr. Paulison is just being demure in not wanting to poke at Congress in describing the true problem here, but if he’s not going to do it who is?
What really catches my eye in the interview, though, is the last question and answer:
Q. Is there a particular disaster scenario that keeps you up at night?
A. What keeps me up at night is a category-4 or 5 coming into this area (New Orleans.) It really does. We could talk about the terrorist issues, with the nuclear bombs, or pandemic flu, but we know we’re going to have hurricanes. We’ve got so many people in travel trailers, so many people in mobile homes, an area that the infrastructure is so fragile. For another category-4 or 5 storm to come in here would be devastating for this entire country. That keeps you awake at night.
There it is. You need no other evidence that FEMA is still fighting the last war. I sincerely hope that FEMA is being a lot more forward thinking than just worrying about another hurricane hit on New Orleans. Just to bring up one example, the next earthquake in LA or SF that rivals the shaking of the 1906 San Francisco quake is projected to do $200 billion in damage, roughly double what Katrina brought to New Orleans.
April 6th, 2007 at 4:38 pm
On NFIP: Can markets for flood insurance function? It seems to me that adverse selection and moral hazard guarantee market failure and indeed NFIP was born precisely because the private flood insurance market collapsed after Betsy. We’re seeing a widespread withdrawal of private insurers from covering coastal properties over enormous parts of the US, so the private market appears once again to be collapsing. So complaining that NFIP doesn’t work like an insurance market is misleading.
There’s clear politicization of FEMA’s flood hazard maps, but there’s also a real problem that recurrence times are nonstationary. A hundred-year flood is not a well-defined concept on a flood plain where changing land-use changes drainage properties on the decadal scale. Under these circumstances, the notion of actuarially sound rates is misleading. This uncertainty seems to be part of the problem for private insurers. Similar uncertainties about earthquake risks seem to be killing the private earthquake insurance market in California.
Finally, on setting priorities: By my reckoning, of the 13 costliest natural disasters in US history (inflation adjusted), 11 were hurricanes and 2 were earthquakes. In the last 20 years hurricanes have caused well over $200 billion in damage in the US. According to the Insurance Information Institute (http://www.iii.org/media/facts/statsbyissue/earthquakes/), a 1906-scale earthquake in San Francisco today would cause about $108 billion in damage. But even if your $200 billion figure is correct, and if this were a 100-year event, it would pale compared to hurricanes.
If, alternately, you count loss of life instead of property damage, the disparity between earthquakes and hurricanes in the US becomes even more pronounced (Of the 15 deadliest natural disasters in US history, 6 were hurricanes and 1 was an earthquake), so I don’t have a problem with FEMA concentrating on coastal hazards. I only wish it were smarter about taking anticipatory action instead of merely reacting after disaster has struck.
April 7th, 2007 at 4:00 am
The civil engineering community has to clean house. It’s easy to find a civil engineer and pay him handsomely for a few days’ work to declare another sliver along the river bottom outside of the flood plain. We also should eliminate the possibility of insuring areas again that have already had damage payouts from the federal dole. If they can get private insurance, fine. But it is a signal of a declining civilization (aka supporting lack of personal responsibility) to continually pay those leaches who rebuild at locations susceptible to disaster.
April 7th, 2007 at 2:22 pm
Jonathan -
1- “Can markets for flood insurance function?” That’s a question I have on my radar to try to answer and I hope that others are working on it too. I don’t think it’s clear, but I wasn’t suggesting that the Congressional meddling with the rates issue is the silver bullet. The mkt failures you raise are serious issues as is the correlated risk issue and Chivers and Flores wrote an interesting paper on a market failure in information:
http://links.jstor.org/sici?sici=0023-7639(200211)78%3A4%3C515%3AMFIITN%3E2.0.CO%3B2-B
2- You’re about right with the numbers … we have done work adjusting the disaster damage numbers and SF 1906 is the costliest disaster on record (by close to triple), 1994 Northridge is #4 and 1989 Loma Prieta is #14. Another way to look at it, though, is that the two costliest quakes on record did as much damage as the four costliest hurricanes. Total losses since 1900 are about double for hurricanes than quakes, so not a huge difference especially considering the much higher frequency of hurricanes. On an annualized basis hurricanes do 3-4 times as much damage. Point is, hurricanes are more frequent but damage potentials are on par. My nightmare scenario is Seattle getting another M9 Juan de Fuca subduction quake directly under the city. I wonder if Paulison is even aware that Seattle has such high quake risk. I wasn’t chiding Paulison for worrying about hurricanes, I was chiding him for being uncreative.