How extreme heat affects insurance—and how the industry can respond

Insurers are feeling the heat as another record-breaking summer finally cools off. In this episode of Critical Point, Milliman leaders Rich Moyer, Garrett Bradford, and Andi Shah—who have researched the impact of extreme heat from the Middle East to Europe to North Carolina—discuss the ramifications on property, workers’ compensation, health, and other types of insurance. They also talk about data, including available sources for studying heat risk now, and the data they would like to see to help measure the future effects of a warming planet.

Transcript

Announcer: This podcast is intended solely for educational purposes and presents information of a general nature. It is not intended to guide or determine any specific individual situation, and persons should consult qualified professionals before taking specific action. The views expressed in this podcast are those of the speakers and not those of Milliman.

Rebecca Driskill: Hello and welcome to Critical Point brought to you by Milliman. I'm Rebecca Driskill and I'll be your host today. In this episode of Critical Point we're going to be talking about extreme heat and insurance.

We're recording this in mid-July, when already the record for the planet's hottest day ever has been broken multiple times, and we're in the middle of an extreme heat wave in the Western and Southern United States. In fact, according to the New York Times, about 24% of the population in the U.S. is forecast to have dangerous levels of heat just today, while Southern and Eastern Europe are bracing for record temperatures that could reach upwards of 40 degrees Celsius or higher. Unsurprisingly, this can affect our health, our infrastructure, and economies around the world.

Today we'll talk about the short- and long-term impacts of extreme heat and how the insurance industry can prepare for rising temperatures. To discuss this, I'm joined by three Milliman experts who have been doing research on the impacts of extreme heat. Rich Moyer is chief product officer for Milliman MedInsight, a platform for healthcare data warehousing and analytics. He's based in Seattle. Garrett Bradford is a geographic information systems consultant based in California. And Andi Shah is a consultant in London who leads our annual study of extreme weather events in Europe.

Rich Moyer: Morning.

Garrett Bradford: Hello.

Andi Shah: Hi, Rebecca. Glad to be here.

Rebecca Driskill: All right. Well thank you all for joining me, and let's just jump right in. I want to start by explaining what we mean by extreme heat. So Garrett, I was hoping you could take this one.

What we mean by extreme heat and how we measure it

Garrett Bradford: Yeah. When we talk about extreme heat, certainly I think we are usually talking about the hot temperatures, the headlines you see, like I'm reading today about Phoenix, Arizona, and temperatures reaching 115 degrees Fahrenheit multiple days in a row.

But it's not always just the hot temperatures, right? Different places experience heat differently. And so one way we often think about extreme heat are kind of locally defined thresholds. So you look at what's your climatic normal over a 20-, 30-year period for a specific place, and we can define our extreme heat events as days or exceedance above some 90% or 95% threshold over that term.

And oftentimes it's not really the maximum temperature that we're interested in. You could look at different metrics like the heat index, for example, also known as apparent temperature. This is a measure of how hot it really feels, combining air temperature and relative humidity. Another metric that's often considered is what's called Wet Bulb Globe Temperature, and that's really the best indicator of heat stress on the human body. It combines temperature, humidity, also wind and solar radiation.

One other metric that we often consider is kind of surprising: it's the minimum temperature, right? On days when our nighttime temperatures don't cool down, the human body is unable to cool itself down, resulting in major health risks. So we often look at minimum temperature as well, as kind of one of the variables used for measuring extreme heat.

The location and length of a heat wave matter, too

Rebecca Driskill: That's interesting, is there anything about longevity, like a heat wave lasting three days versus a week?

Garrett Bradford: Yeah, when we look at these heat events, the intensity is important but also the duration is as well. And going back to that minimum temperature, when you have an extreme event where the nighttime temperatures aren't dropping below a critical threshold where the human body can cool itself off, that's where you start to see really bad health outcomes occurring. So in the case of Phoenix right now, nighttime temperatures aren't dropping below 85, 90 degrees Fahrenheit, and when your body's exposed to that and cannot cool itself down over the nocturnal period over multiple days at a time, you’re much more likely to experience heat stroke.

Rich Moyer: You know one of the things that Garrett talked about that I thought makes it difficult to have a uniform measure of this is it's really location-specific. Being from the Pacific Northwest, extreme heat has a very different meaning than Phoenix. And also the types of mitigation and our ability to cope. Really, you have to know a little bit about the location to know that something that's maybe not that high but a lot of the houses don't have AC, for instance. That's a very different situation than even much hotter temperatures in some place like Phoenix.

Rebecca Driskill: Yeah, and Rich you've done some work in Dubai and in the Middle East, and I imagine that there is a variation in terms of—and Andi actually this is for you, too, in Europe—places with air conditioning, or we see some real geographic variation when it comes to our ability to manage the impacts of extreme heat.

Andi Shah: Yeah, that's definitely something we're seeing in Europe as well. So Southern Europe has typically been warmer than Northern Europe, just location-based. But actually the past few years we've been seeing extreme heat across all of Europe, and southern countries have obviously been affected worse. So, for example, Rebecca, you mentioned this year's heat wave, which we're currently in. For example, in Italy the temperatures are forecast to reach 48 degrees Celsius. That's more akin to what you'd expect in the Middle East, not somewhere in Europe. So again the infrastructure isn't really designed to cope with these sort of extremes, and that kind of ties in with what Rich was saying about location-based, some places will be more capable of coping and others just won't. So for example in the UK, most places don't have air conditioning, and that obviously has an impact on health and infrastructure as well.

The impact of extreme heat on health and health insurance

Rebecca Driskill: Yeah, actually I wanted to talk a little bit about health. So Rich, you read about in the news of course the impact to vulnerable communities, seniors, people whose jobs mean that they have to work outside. What have you been seeing in your research when it comes to extreme heat and correlations in health insurance or the impact on health insurance?

Rich Moyer: Yeah, I think first off there's a lot of work to do that remains. I think measuring the impacts of heat on mortality and morbidity, I think it's relatively new. And we have kind of anecdotal evidence, we have better evidence around mortality. So when people die and it's from heat, there's some pretty good statistics on that. When you start to look at morbidity, which is the burden of illness, we do see examples of where especially cardiac conditions, pulmonary conditions increase during the heat, but we really don't have a sense of the overall impact. And we know the trend is not going well, but we don't know what the real trend and the real impact on overall cost is.

The other thing we're seeing is subpopulations definitely get hurt a lot more than other populations. So the Middle East is a good example. People that are in the air conditioning, that are working in the offices, they're fine. It's the people working outdoors, we have huge mortality for those groups. So it really hits, even within a location, there's really huge differences in populations and how they experience that heat.

Unexpected effects of extreme heat

Rebecca Driskill: That makes a lot of sense, Rich. I was actually just reading an article in the Washington Post about extreme heat and some of the ramifications for employees in the workplace, and I think that that just kind of highlights the systemic impact of extreme heat. We see it on agriculture, it can be city infrastructure as we've talked about, tourism, this is a huge time for tourism in Europe, for instance. It definitely can have trickle-down effects to the economy. So I was hoping that everybody could talk a little bit about some of the lesser-known or more surprising effects of extreme heat.

Garrett Bradford: One of the impacts that always surprised me was the potential impacts to mass transit from excessive heat events. I mean you'd see things like damages to train tracks, roads, and bridges as a result of high temperatures, and then obviously you can't have your safety crews out there for as many hours in the day, so shortening their workdays, certainly impacting their health, and then also just decreased ridership, and what does that mean? So are more people driving on hot days?

And it just kind of really highlights how a lot of our infrastructure isn't necessarily built to withstand the climate change. So we build our cities with one thing in mind, and then over time as our climate changes we need to adapt, and can we do it fast enough?

What happens when buildings and cities aren’t built for extreme heat?

Rich Moyer: Yeah, I think a good example of that was the problems Texas had last year around the power grid. So one of the mitigation ways of handling extreme heat is obviously air conditioning, cooling rooms, and cooling centers, and all that goes away if you don't have a power grid that is functioning. So yeah, being able to keep up our infrastructure and make sure that we have the resilience to mitigate some of these problems, that's going to be a real problem in the future.

Garrett Bradford: Yeah. And one thing that has been of concern in major urban areas is the urban heat island effect. And going back to minimum temperatures and nighttime temperatures, that's where we see one of the most dominant temperature effects right now from climate change is summer nighttime temperatures are rising faster than other temperature metrics. And when you think about urban areas with a lot of impermeable dark surfaces, they're staying a lot warmer throughout the night, causing people to need to run air conditioning later and later and throughout the night in some cases during these extreme heat events, placing even more stress on our grids.

Andi Shah: One of the things that surprised me when I was researching was that actually there's 20% of existing infrastructure in the UK is at risk of overheating, and that includes rail, water, and electricity. And obviously this is going to increase as the temperatures continue to rise. That will obviously have a significant impact on public life. It was surprising. I didn't think it would be that high, but the statistics are there.

How extreme heat affects workers—and workers’ compensation insurance

Rebecca Driskill: And then you get into considerations around business interruption, if there's a rolling blackout, let's say, or you're trying to manage the infrastructure around energy. How does that disrupt businesses and what are the insurance impacts there? Workers' compensation, if someone's more likely to get injured on the job? You get these sort of secondary and tertiary effects, I imagine.

Garrett Bradford: Yeah, certainly. We recently did a study around workers' compensation claims, and even at an annualized level you can see a relationship between indemnity losses for heat-exposed workers and non-. I think that's particularly interesting that some other studies have pointed out. It's not necessarily claims directly related to heat, it's just when you are out in the heat you can be more likely to make a mistake, you can be more likely to slip. So there's a lot of signals in the data that when you look at it might not be claims directly related to heat, but they're occurring at the same time as these major heat events.

Data needed to measure extreme heat’s many effects for insurers

Rebecca Driskill: Yeah that leads me to my next question, which is really how do you begin to measure some of these impacts or future projections? What kind of data do you need when it comes to correlations with health, correlations with workplace safety or business interruption? What data do you look for and where do you get it?

Garrett Bradford: I can start and I'll just start because it's the easiest piece. I mean we can get the geographic data and we can get the temperature data, the humidity data we need for this, at fairly granular scales at long enough time periods to get locally defined thresholds of extreme heat events. So in the U.S. or globally, there's networks of weather station data that you can start with to create interpolated surfaces. There's also a lot of gridded products available. So the geography and the climatology piece is relatively straightforward and easy to get.

I think where it gets trickier is where can we get the claims information, the insurance information to model what's going on? Because we know that things are getting warmer, we know these impacts are going to increase, but unlike some other climate perils we actually have a wealth of data already available, which we can begin modeling and making extrapolations of what to expect in the future.

Rich Moyer: Yeah and I would add, on the claims side, there's a lot of claims data, a lot of clinical data out there, it's getting it in a timely manner, getting it at a level of detail that allows you to be precise, so being able to take a person and geographically assign him to a place. A lot of privacy restrictions make it difficult to put those data pieces together.

The other is that the populations that are the most vulnerable also of course are the ones with the least amount of information. So while employed people, Medicare, we have great data. If you're homeless and uninsured, that's really difficult data to get to. So I think we've got good data assets, we're making a lot of use of it, we're learning a lot from it, but there's definitely some gaps and some things that would make the data more usable, especially timely. I think especially with temperatures changing so fast, having studies that are two years behind, just it's not keeping up fast enough. So we really need to accelerate the pace of things.

Andi Shah: Yeah, I agree. It's very difficult to tally the monetary losses that come from heat and drought because effects like the industrial slowdowns due to lack of cooling water, those things take time to emerge. It's not like floods, where the damage is more visible and therefore can be assessed quicker. So I agree timeliness is definitely an issue.

One of the things the EU has done recently is their climate monitoring service, Copernicus, has got open-source data that's free and accessible to all. It is mostly satellite imagery, but that sort of data can be used to perform analysis. I take your point, Rich, that it is probably a bit out of date, but it's something that some of our European colleagues have already been using to do research and help their clients. So for example our French colleagues, they have used sort of before-and-after satellite imagery to analyze the severity of wildfires and map that to areas in France which could be affected. So there is analysis that can be done, but as I mentioned it's a little bit difficult to get timely information particularly from extreme heat.

Types of heat to consider when assessing the risks of high temperatures

Rebecca Driskill: Thanks, Andi. I think that that really nicely leads into my last question, which is what or how can insurers prepare? What can they be doing to help manage the risks of extreme heat? And then I'm also curious from you all if you could wave a magic wand and have all the data that you needed, what would you like to be modeling, what would you like to be looking out for?

Andi Shah: I think when insurers look at high temperatures as a risk, they should consider two distinct types of heat. So either drought heat or humid heat. Drought heat is the one that leads to water deficit and can kill trees and crops, shrink clay soils, and displace shallow building foundations. So those are the sort of things that could lead to subsidence claims coming through. Drought and heat together also significantly increase wildfire risk.

When we look at humid heat, this is what occurs when high humidity accompanies high temperatures, and this is the most dangerous to humans because it compromises the ability to perspire and have a cooling effect. Humid heat may increasingly be covered by cancellation insurance for outdoor events when temperatures are dangerously high. And obviously there's an impact on life and health insurers as well.

In terms of the second question that you asked, Rebecca, I guess to answer your question about what magic data or what magic analysis we could do, it would really be to look at the data that's available in terms of satellite imagery but also try and supplement that with things that people have done that may not be captured by the satellite imagery. So, for example, when I look at July 2021, where London particularly had a lot of flash flooding that took place, most of that happened because basements got flooded and basements were not really captured in any of the modeling that was carried out. So that sort of risk wasn't being captured. And I guess that's an extreme of measures not being taken into account. But you could flip it around and say that some policyholders have done stuff maybe to their property, which can help mitigate any kind of subsidence risks, for example, and those sorts of things would need to be captured or ideally should be captured.

Using population health methods to study the payer impact of extreme heat

Rich Moyer: I can talk about it on the health side. One is I think insurers need to work with communities and providers, and there's some good examples, not in the climate area as much yet, but I think there's good examples of insurers working with communities and providers, so that's working together.

Second is to create the data assets that are needed. So in addition to the geographic information and the information about heat and smoke and other climate impacts, but also making sure that you have full clinical information in a timely manner and that you could connect those pieces at a precise enough level to know what the impacts are.

And then I think the final thing is really treating this like other populations. So, if I have a population of diabetics, I want to use population health concepts, which is, "Here's the things I know about this population. Here are the risks. Here are the subpopulations that are at disproportionate risk." And then having strategies and tactics to mitigate those. So in the heat context it's really, "OK, I know who my population is, I know which segments of the population are severely impacted. So people who work outdoors that have high cholesterol, that may exacerbate their chronic conditions, let's target both as a community and as an insurer on those." And that's population health, that's what the industry's been moving to. So I think this fits in nicely with—but thinking of it as a unique, specific population that has its own needs, and then working with the community to work on those needs.

Extreme heat affects all types of insurance

Garrett Bradford: Yeah, what I've been saying for a while now is, whether or not insurers are modeling this or incorporating excessive heat, it affects their portfolios. We just talked about how the impacts of extreme heat are systemic across the economy. We see the effects on gross domestic product (GDP). So it is affecting every sector.

So what should insurers be doing? I think they should be looking at what has been the historical impact of excessive heat on their policyholders, on their claims, and use that to take action. If it's on workers' compensation, requiring your employers to educate their workforce or have responsible workplace heat standards. These are real steps that could mitigate claims and future losses.

On the health side, we know that there's ways that we can mitigate our urban environments to reduce heat stress on populations. So can we find creative ways to fund those projects, knowing that by mitigating urban heat we could potentially reduce utilization or claims?

And then to answer the question about the magic wand, I mean obviously Rich alluded to some of the trouble in getting kind of granular claims information. I mean, being able to tie claims to specific times and place is critical to modeling this stuff. I'm a geographic information systems (GIS) nerd, so I want to know exactly where these happened. Ideally the exact latitude and longitude, but if we can't get that, well can we get down to the census block, can we do this in a way that's going to protect the health information of the individual while still being able to take some meaningful action away from this data?

Rebecca Driskill: Thanks. I think those are all great points. I wish I could wave my magic wand but, in the meantime, we'll just continue to do our research. And I want to thank my guests Andi, Garrett, and Rich for joining me today. You can read our latest climate resilience research, including our “Extreme Weather Events in Europe” paper, on Milliman.com.