The burning of California

How did this research come about? How do wildfires factor into your work at the Risk Center?

Kunreuther: The research first started with a project that the Risk Center did with [California utility company] Edison International that focused on the cost of wildfires to the utilities, to the state, to insurers, and those suffering damage. We explored the question: Who is responsible for the cost of the wildfires? Erin and I have now been focusing on the behavioral aspects: Why people do not protect themselves and their homes against wildfires, what communities can do to reduce those losses, the challenges we face given the cognitive biases that impair sound decision-making, and how we can encourage and incentivize the community as well as individuals to take protective actions before the next disaster occurs. Right now is the time to examine ways to reduce the wildfire risk because people are concerned about losses from wildfires.

St. Peter: The Risk Center has been looking at natural disaster risk and insurance since its founding. We’ve long been thinking about climate change, but are now seeing increasing problems managing wildfires and insuring in wildfire-prone areas. Faculty and staff at the Risk Center have therefore started looking more closely into these types of perils.

What are some of the causes of the increasing severity of wildfires in California?

Kunreuther: Ninety percent of U.S. wildfires are caused by either technical sources, like utility wires falling on trees—that’s why the utility companies have been concerned—or by humans, like leaving campfires unattended. Notably, almost all of the current wildfires in California were caused by lightning strikes, and that’s very different from what triggered wildfires in the past. The impact of the current fires are extensive. If you compare 2019 to 2020, from Jan. 1 to Sept. 15, 2019, 4,129 fires in California had burned through 43,417 acres. Contrast that with the period of Jan. 1 to Sept. 15, 2020, more than 6,352 fires had chewed through 1.4 million acres. So you’re really talking about significant wildfire damage and losses. Climate change is going to make wildfires worse in the future due to more droughts and severe heat waves.

St. Peter: The combination of these different types of causes makes it really difficult for insurers to model wildfire risk and even price wildfire insurance. Lightning is a little bit easier to model, but the fact that someone can start a wildfire anywhere makes it very difficult for insurers to figure out how to assess the underlying risk. You don’t have that same human element in, say, hurricanes. Humans can’t cause hurricanes—at least not in a direct manner in the way that they can cause wildfires. The immediate causes are really interesting and different when you compare them to other natural disaster perils.

Why aren’t Californians and California governments doing more to protect themselves and communities from wildfires?

Kunreuther: This, of course, is what we’ve been researching. Most people underprepare for these disasters. They don’t take the right steps in making their houses safer. We often assume wildfires and other natural disasters will not happen to us. There is a tendency to be myopic so we don’t focus on long-term strategies. We are optimistic and underestimate the likelihood of a wildfire occurring. We tend to follow along with what our friends and neighbors do—but they are unlikely to be more informed than we are. Communities can play an important role in reducing the wildfire risk through well-enforced building codes and brush clearance regulations, such as requiring residents to clear vegetation within 100 feet of their homes to reduce the spread of these fires. 

We feel it’s extremely important for those at risk to recognize that they need to focus on the impact that wildfires can have on their property so they undertake loss reduction measures now. I’ll give you an example that we use in the blog that Erin and I wrote, ‘Reducing California Wildfire Losses.’ When you tell a person there’s a 1 in 100 chance that there’s a wildfire that might cause damage to their house next year, what do you think they say?

Probably that it’s a low risk.

Kunreuther: Yes, they think, ‘It’s not going to happen to me. I don’t worry about it.’ But if you tell them that there’s a greater than 1 in 5 chance of at least one wildfire causing damage to their house in the next 25 years, they respond, ‘1 in 5?’ That doesn’t sound like a low risk. And they pay attention. But that probability is the same as an annual probability of 1 in 100 if wildfires happen randomly each year. We are framing the likelihood of a wildfire in a way that gets those at risk to recognize what could happen to them if they haven’t protected their property. The community is really important because it isn’t just one home that has to protect itself, all the homes in the community need to do so because fires are an interdependent risk. So it is important for communities to have well-enforced standards and zoning regulations to reduce the likelihood and consequences of wildfire risk.

St. Peter: There are certainly a lot of behavioral elements at play that apply to other types of natural disasters, and Howard is the longtime expert in the field of behavioral economics. Besides the behavioral elements, there are regulatory barriers to change, too. The regulation of the insurance industry in wildfire-prone areas is an example of this. In some states, the insurance departments play a role in regulating premiums for different insurance policies. In California in particular, the insurance regulators do not permit insurers to use catastrophic risk models or to include the cost of reinsurance in rate-setting. Insurers operating in California are thus only able to use prior historical data on wildfires to price their premiums. Since insurance prices should convey information on the underlying risk, these limitations restrict the ability of premiums to reflect rapidly evolving risk levels and may cause individuals to underestimate their exposure to wildfire risk.

Are there things that individuals could be doing that they aren’t doing?

Kunreuther: Absolutely. They can make their house safer by putting screens over vents and replacing siding and roofs with less flammable materials to reduce the likelihood that embers will create a wildfire. They can take care of their backyard and make sure that they don’t have a lot of wild grass and vegetation that would spread the wildfire to other homes. It’s related to the interdependency of the risk. If each of the individuals do that, then there’s a good chance that the community will not have a major fire. Even if lightning strikes one home, fire may not spread to other homes.

St. Peter: In Boulder County, Colorado, there’s a really neat program that matches residents with professional inspection resources to get a home inspection with a detailed strategy that allows them to know exactly what they can do to lower their risk. Such efforts become a lot more powerful when they’re done at the community level.

When you talk about interdependency, do you mean everybody taking collective responsibility to try to prevent wildfires?

Kunreuther: That is part of what I mean; interdependency also implies that even if you have undertaken protective measures, you may not be safe if your neighbor hasn’t followed suit. It’s an interdependent risk in the sense that if your neighbor has decided not to invest in loss reduction measures and you took steps to protect your home, you could still incur significant damage because a severe fire to your neighbor’s property can spread to yours. Some homeowners in high-risk areas have indicated that they feel it is not worth it to invest in protection if their neighbors are unconcerned about suffering damage from wildfires and have ignored the risk. That’s the importance of a community playing a role.

From my perspective, as someone who’s in Pennsylvania, it seems like California has massive wildfires every year.

Kunreuther: It’s happened a lot, you’re absolutely right. Colorado and California are two states that have suffered significant losses. And Oregon and Washington are now experiencing severe damage from wildfires.

But it sounds like some Californians aren’t necessarily taking the threat seriously even though it seems like it’s an annual threat. Why is that?

Kunreuther: Well, that’s the point that Erin and I are making with respect to these biases. If your house hasn’t suffered damage from a previous wildfire, you don’t think it’s a problem. A lot of these people have not had a personal experience with wildfires, so they think that they’re safe, or they think that the likelihood of a wildfire damaging their home is so small that they don’t want to pay attention to the potential consequences. That is a major problem.

St. Peter: While it might be an annual thing for California, not every neighborhood or every county is getting hit each year, so there’s certainly that question of how much information stays in your working memory. You might be a new homeowner in California. You might notice that your particular neighborhood hasn’t had a fire in 10 years and maybe you think you’re safe. I think it’s important for individuals, governments, and nonprofits to help residents of California realize that just because something hasn’t happened in your neighborhood in the last 10 years doesn’t mean that that’s the true probability of your underlying risk. These are oftentimes low-probability catastrophic events. They’re becoming higher probability events.

With floods, there are flood zones. With earthquakes, there are fault lines. With hurricanes, there is Hurricane Alley. Does a similar danger zone exist for wildfires?

St. Peter: You don’t have the well-known flood hazard boundary for wildfires that you have in flood insurance. And with flood insurance, the U.S. is moving toward better disclosure of flood hazards via federal mandates and local outreach activities enacted by local governments participating in the Community Rating System. A federal mandate requires mortgage lenders to disclose to borrowers that a home is in the 100-year flood plain and requires them to purchase flood insurance. Some related efforts do exist for wildfires, but they can be more localized. For example, the California Department of Forestry and Fire Protection, CalFire, does identify and publish ‘fire hazard severity zones’ and mandates that sellers disclose whether a property is located in the very high wildfire risk zone during residential property transfers.

Does California have any universal fire prevention codes or do anything to promote more awareness? And is this more of a local problem and solution or a statewide problem and solution?

Kunreuther: The state and many communities are promoting awareness of the risk. Communities are responsible for developing and enforcing building codes and zoning regulations. Often, a community wants to increase development and so they don’t have a lot of these restrictions. It is really important for them to have inspections of property and enforcement of building codes. It isn’t just a state problem, it’s a local problem.

St. Peter: There are things that can be done at the state level to re-evaluate how different regulations incentivize development in high-risk areas, but there is a lot of local control as well. So many of the zoning decisions are made at the municipal level so local governments must be part of the solution. Even though wildfires happen every year, these recent fires have been very big, so the hope is that there will be a rallying call that gets both state and municipal agencies coming together to develop some strategies moving forward.

Are you seeing any improvements or a heightened awareness from California considering what they’re going through, and do you think the current catastrophe will bring about change?

Kunreuther: As Erin and I note in our blog, a zoning reform bill was introduced in the California state legislature at the end of 2018 encouraging housing growth in low-risk areas like San Francisco and Los Angeles—urban areas that don’t have a wildfire problem—to reduce the pressure to develop in the wildland urban-interface, the high-risk wildfire areas. It normally takes a wildfire like the current ones to spur new legislation because the public demands it. Normally it is only when disasters happen that people and legislators pay attention.

St. Peter: In the past week or so, it has become part of the rhetoric in the upcoming presidential election as well. While these wildfires are extremely unfortunate and disastrous for the communities that they’re affecting, I am hopeful that the increased media attention might allow for change to start happening. But these are very complicated issues. Wildfire risk is not the only thing that municipalities are worried about when they’re doing their zoning plans and trying to come up with tax revenue. It is hard to make improvements on the issue given that doing so might also step on a lot of toes and interests in regards to land-use planning.

Since your blog post was published, large wildfires have also affected Oregon and Washington. Do you think your research and findings apply to those states as well?

St. Peter: A lot of the research is generally applicable. Some of the solutions on how to regulate the insurer-side of the wildfire problem may look different from state to state because of how insurance is regulated at the state level. But I do think that certainly the behavioral elements of why we underprepare for disasters, and the challenge between wanting insurance to be affordable and the need for premiums to reflect underlying risk, are going to be the same problems no matter where they’re happening. The particular solutions or the particular movers and shakers might be different in how you enact change, but I think the broad themes can definitely be applied to those areas as well.