Based on the notion that sunlight is the best disinfectant, I've long advocated that homeowners insurance companies give clients as much information as possible about the risks they face. Don't just quote me a premium. Tell me that, perhaps, I'm at more risk of flood or wildfire than old government maps show--and help me understand what I can do to reduce those risks.
Zillow just took a step in the opposite direction.
It had announced 15 months ago that it would feature detailed climate risk information for flood, wildfire, wind, heat and air quality, but the company quietly dropped that information last month.
The reason is obvious: pressure from sellers who didn't want the risks to their properties spelled out.
The implications are disheartening.
What Zillow was attempting was always going to be tough, because we humans aren't wired to think rationally about probabilities. If some political poll says a candidate has only a one-in-10 chance of winning, and they win, we leap to the conclusion that the poll was wrong and the pollster incompetent. Maybe. But maybe not.
The only way to test is to look over a body of work and over time. Did those predicted to have a one-in-two chance win about half the time? Did the one-in-fours win a quarter of the time? Did those one-in-10s win a tenth of the time?
But models like the one from First Street that Zillow used haven't been around long enough for us to have much evidence about whether they're right when they say there's a one-in-50 chance of a wildfire affecting a home each year.
A spokesman for First Street said, "During the Los Angeles wildfires, our maps identified over 90 percent of the homes that ultimately burned as being at severe or extreme risk — our highest risk rating — and 100 percent as having some level of risk, significantly outperforming CalFire’s official state hazard maps.”
But we humans are still wired to think, "Zillow said I was at severe risk of flooding, and I didn't have a flood this year, so those bozos were wrong." In the context of the risk ratings provided by Zillow, someone with a house to sell would surely also think, "And their error is costing me money."
While that sort of thinking led to enough pressure on realtors, a key constituency of Zillow's, that Zillow pulled the ratings, there's still some hope for the long run. Even Zillow still provides a link to First Street so those curious enough can find information about risks to properties they might buy. And good models like First Street's will not only get better but will be more accepted over time, as they build up a track record.
It'll just take longer than I had hoped, perhaps much longer.
Sorry, I don't make the rules. I sure wish I did....
Cheers,
Paul
P.S. So I don't end on a total downer, I'll share two links that contain a healthy dose of encouragement. First is a webinar I did recently with Francis Bouchard, a managing director at Marsh McLennan who has focused on resilience for years, and Nancy Watkins, a principal at Milliman who has developed a Data Commons to help mitigate wildfire risks in the wildland-urban interface. Second is the ITL Focus from September on resilience and sustainability, featuring an interview with Francis and parts of an interview with Nancy.
Both describe the sort of conversation that insurers need to have--and are starting to have-- with architects, builders, city planners and others so that, as a group, we can build resilience into properties from the outset and can at least offer advice to homeowners and communities on how to reduce risks related to severe weather.