Wildfire Risk and Home Prices, with Lala Ma

Notable quotes

• Helping homebuyers account for wildfire risk: “Damages from wildfires have increased in recent years, so it’s really important that we know whether people are internalizing these risks. If they’re not, it suggests a potential role for public policy to help facilitate this information—in this case, through public disclosure.” (6:28)
• Disclosing wildfire risk reduces home prices: “Houses in regulated areas … sell for about a 4 percent discount compared to those in nonregulated areas … You might interpret this as, disclosing the wildfire risk causes prices to fall by 4 percent, on average.” (14:23)
• Income level influences willingness to pay to avoid wildfire risk: “Higher-income households have higher willingness to pay to avoid wildfire risk. They have to be compensated more to contend with this higher risk. This is regardless of whether disclosure is regulated or not in an area.” (19:50)

Top of the Stack

• “Risk Disclosure and Home Prices: Evidence from California Wildfire Hazard Zones” by Lala Ma, Margaret A. Walls, Matthew Wibbenmeyer, and Connor Lennon
• Books by Emily Oster, including Expecting Better and Cribsheet
• The Two-Parent Privilege: How Americans Stopped Getting Married and Started Falling Behind by Melissa S. Kearney

The Full Transcript

Margaret Walls: Hello, and welcome to Resources Radio, a weekly podcast from Resources for the Future. I'm your host, Margaret Walls. My guest today is Lala Ma. Lala is an associate professor of economics in the Gatton College of Business and Economics at the University of Kentucky, where she holds the Carl F. Pollard Professorship in Health Economics. Lala is, I might add, a new addition to the family of university fellows at Resources for the Future (RFF), which I'm really pleased about.

Lala's research focuses on estimating values of environmental quality as assessed through housing markets, and some of this research involves use of what are known as “hedonic property value methods,” which we're going to talk about today, and she's also an expert in the estimation of what are called “locational-sorting models,” models that look at the factors, including environmental factors, that determine where people live. A lot of Lala's work examines a distribution of outcomes across people of different incomes, and she's written several papers on environmental justice.

Today, we're going to talk specifically about a paper she recently published with coauthors in the journal Land Economics. This paper just came out and it's titled "Risk Disclosure and Home Prices: Evidence from California Wildfire Hazard Zones." Now, wildfire is a serious problem in the western United States, and we've had other podcast episodes about wildfire, so many of our listeners probably know about that. But one of the key reasons for it is population growth in what's called the “wildland-urban interface,” or the WUI. It's an open question whether and to what extent people really understand the risks and take them into account when they make those location decisions. This paper looks at the important role of risk disclosure, and we're going to talk to Lala about that paper, its findings, and also some ongoing work she has on wildfire risks. Stay with us.

Hello, Lala. Welcome to Resources Radio. Thanks so much for coming on the show.

Lala Ma: Thanks for having me, Margaret.

Margaret Walls: Before we dive into our topic today, I want to ask you to tell us a little bit about yourself and how you came to your particular career path and your work on the combination of environmental, housing, health, and equity issues. Can you tell us about that?

Lala Ma: Sure. I was an undergraduate economics major at Tufts in Massachusetts, and what drew me to the major when I was an undergraduate student was just how some of these economic models might explain individual behavior and decisionmaking. I think I did have an early interest in looking at how data might support that. Then, in graduate school, one of my favorite courses was on something called “nonmarket valuation.”

Here, what I thought was really cool was how you can use people's decisions to back out or reveal their preferences for an attribute attached to some good that they were deciding over. The example that's relevant to what we'll talk about today is the housing market. People make decisions about where they want to live, and this is based on attributes that are attached to the house, one of which could be environmental quality. If environmental quality might impact the price, we might think that the contribution of environmental quality to price is an implicit price for what people are willing to pay for that good.

The nice thing is that we can then, by getting an implicit price, try to calculate, for example, the benefits of a policy that affects the environment and then compare that to other policies that might have nothing to do with the environment. In this area, you can get at people's preferences by using housing market analyses or studies, but you could also try to get at valuation in other ways.

You mentioned I had some work looking at health, and one way is to measure the direct impacts on health of exposure to an environmental bad. Then, you can monetize that using something like an estimate of what people are willing to pay to avoid the risk of death or illness. I've done some work in this context, looking at shale gas development.

Then, with respect to the work on equity, I think the more I worked in this area on valuation, it became apparent that people held different values for environmental quality. You can infer this based on data showing that different types of people are exposed to different amounts of pollution and, in particular, along the lines of socioeconomic status. Then, the question is, Is this actually because people had different values for environmental quality? Or is something else going on—for example, other forces, such as regulation or decisions made by other stakeholders that are contributing to disproportionate exposure? All this contributed to my interest in equity issues and environmental justice, as many call it.

Margaret Walls: That's good background for this paper. I gave a little bit of background at the beginning. Tell us a bit more about it, Lala—what exactly you and your coauthors set out to look at in this study and what the motivation was for this work.

Lala Ma: In this study, my coauthors—you, Margaret Walls, Matt Wibbenmeyer, and Connor Lennon—are using purchase decisions in the housing market to try to estimate whether and how disclosure about wildfire hazard, through regulation, impacts housing prices. If housing prices are impacted by disclosure about risks, then we take this as evidence that people's decisions regarding where they lived are being influenced by whether there is information about fire risks and whether it's disclosed.

If we find no impacts, it suggests that risk disclosure, in the wildfire context at least, is not so useful for homeowners, and that people already have full information, or all the information, they need to make their housing choice. As you mentioned earlier, damages from wildfires have increased in recent years, so it's really important that we know whether people are internalizing these risks. If they're not, it suggests a potential role for public policy to help facilitate this information—in this case, through public disclosure.

Margaret Walls: I want to draw a parallel to flooding, Lala, and I know you've done some work on flooding, so there's some similarities and some differences. What are the key differences in mapping and information disclosure of risks when you compare these two hazards?

Lala Ma: That's a great parallel to draw. We think the most important differences between wildfire and flooding are twofold. First, wildfire insurance is generally provided through homeowner insurance policy. It's not a standalone policy that you purchase, whereas in the case of floods, it’s insurance you have to purchase separately. It's possible that people may let their policies lapse.

Second, there's no federally mandated maps on wildfire risks, whereas in the flood case, the National Flood Insurance Program requires mapping of flood risks through flood insurance rate maps. I think those are called FIRMs. These two differences affect how people might trade off costs of living in a location with disaster risks depending on whether they have insurance, their insurance coverage, and the extent to which people even know about the risks that they're potentially facing.

Margaret Walls: Our paper focuses on California, and that's the state with the biggest wildfire problem. It is also the state with some interesting and unique disclosure requirements that made this study possible. Can you tell us about how disclosure works in California?

Lala Ma: Yes, sure. California law requires sellers to disclose the wildfire hazard for their properties in certain areas where there is higher risk. The types of areas that require disclosure are based on the degree of wildfire hazard, and this is known as the “fire hazard severity zone,” and this is delineated by different maps. Then, it also depends on the type of jurisdiction a property is in. There are several types of jurisdictions: federal, state, and local jurisdictions. We call these “responsibility areas.” During the timeframe of our study, all properties located in very high fire hazard severity zones require disclosure when the house is being sold. Then, properties in moderate and high fire hazard severity zones in state responsibility areas also require disclosure. The other areas do not.

Margaret Walls: That's good background, too, because we're going to come back to that. Let me ask you now about the data and the methods that you use. Can you tell us about the data first? What data was involved here?

Lala Ma: We used housing transactions data for the universe of home sales in California. The period of analysis for our study is 2015 to 2022. These data were made available through Zillow’s Transaction and Assessment Database. This is really cool data, because it gives us information on individual home sales. This includes information such as the sale date, buyers and sellers, the characteristics of the houses—the number of bedrooms and number of bathrooms—and, importantly, the location of the house. The precise location of the house allows us to map where houses are relative to these fire hazard severity zones and responsibility areas.

I do want to say, by us mapping—and by us, I mean a very talented research associate at RFF who mapped and recovered these spatial relationships between the homes that were selling and whether they're in one of these disclosure or high risk areas—after we've done this mapping, this allows us to figure out whether a sale occurred in an area where disclosure was required or not.

Margaret Walls: That research associate is Emily Joiner, and she is quite talented, so we'll come back to her, too.

Once you have all this data, there's always some challenges in these kinds of models, because lots of things are correlated with each other. Teasing them apart and identifying the effects of one thing, in our case wildfire risk disclosure, can be hard. Talk about that a little bit, because that requires some special techniques. What did we do to handle that problem?

Lala Ma: This issue of teasing out what contributes to price is prevalent in all sorts of analyses that use this type of data. I should mention that our paper, which is published in Land Economics, is part of a special issue where all the papers in the issue use this Zillow data for different applications. The specific challenge that we're dealing with here is that higher wildfire risk areas are more likely to require disclosure, but then places with higher wildfire risk also have shown that they've had better access to amenities, like greater access to open space, better views, etc. This means that the price difference of disclosure-regulated areas versus unregulated areas not only includes the effect of disclosure on prices, but also the price effects of these differences in other amenities that's correlated with disclosure status. We control for a lot of these differences in our analysis, but it's impossible to control for everything. So, we need a strategy to deal with this.

What we do then is the following. We focus on a very narrow neighborhood around these regulation boundaries—specifically, 300 meters around these regulation boundaries—and we compare the prices of houses sold on the regulated side with the unregulated side. In this narrow neighborhood, when you cross that boundary, disclosure regulation changes discretely, but the amenities, whether we observe them or not, should stay relatively the same in this narrow window. What this does is ensure, or better ensure, this apples-to-apples comparison. In other words, the only difference between houses on either side of this boundary is disclosure regulation, so that any price difference that we capture could be attributed to that difference.

Now, we also don't want this price difference to be due to increasing fire risk, going from the nonregulated to the regulated side, since we're trying to capture that pure effect of disclosure, so what we also do is focus on high fire hazard severity zone areas and also control for a continuous measure of fire risk, which is the wildfire hazard potential.

Margaret Walls: Right. Then we're just picking up the regulation and not the change in the hazard, right?

Lala Ma: That's right, exactly.

Margaret Walls: A lot of hard work goes into this. Tell us about the findings. What did we find in the study?

Lala Ma: We found that houses in regulated areas that are within that 300-meter boundary sell for about a 4 percent discount compared to those in nonregulated areas. I think that's our headline result. If you're willing to believe that we've done a decent job in purging the effects of those other correlated differences with disclosure, you might interpret this as, disclosing the wildfire risk causes prices to fall by 4 percent, on average.

Margaret Walls: Do you interpret that, then, as disclosure making a difference?

Lala Ma: Exactly. It suggests that disclosure is potentially imparting additional information about fire risk to people, and this is causing them to affect what they're willing to pay for a property.

Margaret Walls: People could have information from other sources, so is it a lower bound on that effect? Or what do you think about that?

Lala Ma: For one thing, we are comparing people or purchases on the regulated versus nonregulated side. What we're really capturing is a difference that disclosure makes. It's possible that people in areas where risk is not disclosed still have information about risk. In that sense, we're removing that discount they're already willing to pay. So, this would be considered a lower bound on the actual willingness to pay to avoid wildfire risk.

Margaret Walls: That roughly 4 percent figure—did that magnitude make sense to you, Lala? Is it in line with any other studies you know about that are out there either for wildfire, or, maybe, in the flooding space, or something like that?

Lala Ma: We did a little bit of a literature review in the paper, and, generally, the magnitude seems reasonable based on what previous work has found. There are various strategies that people use to try to isolate the effect of disclosure on prices, or risk on prices, and so many studies in the disaster context … We're looking at regulation in these fire hazard zones—how that changes prices or disclosure—but many of the studies in the disaster context use disastrous events to try to recover the effect of risk on prices.

Now, there are clear differences between that and what we're doing, but you can think about it in a similar way, as if wildfire events are communicating risks to individuals, similar to how disclosure is communicating risks to individuals. We also compare our work to this type of work, and it seems that our estimates are a little bit smaller in magnitude, but they’re in the range.

Margaret Walls: We have some follow-up work going. You can offer a teaser to our audience about this, and this involves the development of a locational-sorting model, which I explained at the beginning, as you're an expert in that, and this is in the wildfire context. Why don't you tell folks what we're doing now on this project and, maybe, we have a few preliminary findings, so stay tuned, folks. Tell everybody what we're working on now. It's kind of an extension.

Lala Ma: In this follow-up work, we're still interested in the effects of disclosure, but we're going at it in a slightly different way, and two main things are different now. First, as you mentioned, we're explicitly modeling how people choose where to live, and this is through the sorting model. Then, second, in this follow-up study, we’ve recovered more information about the individual homeowner making that housing decision, whereas in the previous paper, we're really only looking at detailed information about the house rather than the person buying the house. Combined, these two things allow us to better explore heterogeneity and preferences to avoid wildfire risk.

We plan, or have started to look at, how demand to avoid wildfire risk will vary by disclosure regulation and by income. This might be interesting to look at, because it might raise some equity issues with policies that try to either address wildfire risk directly or indirectly through providing information, for example, through disclosure. We've also mapped out a different form of heterogeneity. We've mapped out a person's wildfire experience. This is whether a wildfire occurred near a person's prior residence before buying their current home, and this allows us to think of a different type of risk-information communication, like what I mentioned earlier, so we're interested in seeing how this might matter in housing decisions with respect to wildfire risk.

If you want a teaser, I don't have specific numbers for you, but we're finding some interesting patterns. Currently, we're finding that higher-income households have higher willingness to pay to avoid wildfire risk. They have to be compensated more to contend with this higher risk. This is regardless of whether disclosure is regulated or not in an area. We're also finding, like before, that disclosure matters. In areas where disclosure is not regulated, the willingness to pay for wildfire risk is positive. This is a little bit counterintuitive, but then, when you look at disclosure-regulated areas, this willingness to pay becomes negative and statistically significant. Disclosure is making a big difference there. Then, interestingly, on the fire-experience part, we're seeing that people with fire experience in the past actually have lower willingness to pay to avoid wildfire risk. We're still trying to interpret that and to figure out what's going on, but it does suggest, if anything, there's an important distinction between the source of information of these risks. So, is it experience, versus ex ante risk disclosure?

Margaret Walls: Do you want to just say for a second, the data issues—what did we have to do to both get these individual characteristics and then match them with whether they had fire experience? Say a little bit about that.

Lala Ma: This was a challenge, and I'd say the other members of our team did a fantastic job in doing this. For recovering information about the individual, we had income information and information about prior fire experience. For income information, we merged these housing-transactions data with data from the Home Mortgage Disclosure Act. This is basically mortgage-applications data. We did this merge looking at information on the year of sale, the census tract of where the house was sold, and the loan amount. Those things are available in both data sets. If we can match, in this mortgage-applications information, we can get a self-reported measure of income and also race, if you're going to look at that, as well. That's how we got income information.

The second piece, which is this fire experience—the Zillow database, which is the housing-transactions database I mentioned, had individual buyer and seller information. What we did also was look at the name of the buyer and look within a year of that transaction whether, within a year, there was a seller of the same name, and we looked at that location of where that seller sold their property. We inferred from that where the person was located previous to their current purchase, and then, based on that location, we mapped it to fire events that occurred, which allowed us to infer whether there were any fires in that area.

Margaret Walls: A lot of work went into that. Neither you nor I did that part of the work.

Lala Ma: No.

Margaret Walls: I think there's going to be some interesting findings here based on both the income information and the fire experience. There's some new insights. One thing we've talked about, Lala, is policy simulations, like, What could we look at with this model? Those are things we haven't done yet, but maybe you want to say a little bit about that. What are the advantages of doing a model like this for looking forward?

Lala Ma: Based on some of these modeling assumptions, you can back out a welfare measure. How does a change in some particular dimension make someone better off or worse off, given their choices? One of the things we had talked about potentially looking at is, Suppose we change the cost of insurance in these areas, does this make people worse off? If it's more expensive, it probably makes people worse off.

But the question of interest also is, How does that welfare change vary by income—in this case, the income gradient? That's something we can calculate and look to better understand the distributional consequences of insurance reform, for example. The other thing I think would be really interesting to look at is this value of information, because we are getting this value of disclosure or the impacts of disclosing wildfire risk. We could look at, in the absence of disclosure, How are people made worse off? I think those would be valuable counterfactuals to look at with this model.

Margaret Walls: I do, too. The insurance question looms large these days; that's for sure.

We're running out of time here, and we need to close our podcast with our regular feature, which we call Top of the Stack, and that's where I would like to ask you to recommend a book, an article, a podcast—really anything at all that's caught your attention lately. What's on the top of your stack?

Lala Ma: I'm about due for my second child, so I've been reading a lot of pregnancy books, and this is not a new book, but I highly recommend the series of Emily Oster's books, just for an economist … It's a great literature review of everything you might need to know for pregnancy. That's been taking up a lot of my attention.

But I recently picked up a book called The Two-Parent Privilege. I don't know if you've heard of this book. This is by Melissa Kearney, who's an economist at Maryland. This book is about the economic impacts or implications of marriage. I think she provides lots of data-based trends of marriage in the United States and their aggregate impacts. I haven't gotten very far in the book, but a lot of my interests here came from my own upbringing in a single-parent household, and it's really hard to think about the counterfactual of how I would've turned out if that hadn't been the case. But it does make me reflect back on how policy might have influenced my family structure, the family structure I was raised in, and how things could have been very different. It also makes me feel very grateful that I have someone else I can rely on to help watch my kid if someone gets sick.

Margaret Walls: I think I've heard of that book. That's a really good point. Well, those sound great. Thank you so much for sharing. It's been a pleasure having you on Resources Radio, Lala, and learning more about your work, our work. We're going to stay tuned for the new research and thanks so much for taking the time to come on the show.

Lala Ma: It's great to talk about this and talk with you. Thanks.

Margaret Walls: You've been listening to Resources Radio, a podcast from Resources for the Future, or RFF. If you have a minute, we'd really appreciate you leaving us a rating or a comment on your podcast platform of choice. Also, feel free to send us your suggestions for future episodes.

This podcast is made possible with the generous financial support of our listeners. You can help us continue producing these kinds of discussions on the topics that you care about by making a donation to Resources for the Future online at rff.org/donate.

RFF is an independent, nonprofit research institution in Washington, DC. Our mission is to improve environmental, energy, and natural resource decisions through impartial economic research and policy engagement. The views expressed on this podcast are solely those of the podcast guests and may differ from those of RFF experts, its officers, or its directors. RFF does not take positions on specific legislative proposals.

Resources Radio is produced by Elizabeth Wason, with music by Daniel Raimi. Join us next week for another episode.