A recent study by Philip Mulder with coauthor Benjamin Keys of The Wharton School finds that the need for disaster risk reinsurance—financial backup insurers purchase from global reinsurers to avoid paying out more in claims than they collect in premiums—is a key driver behind the rising cost of homeowners insurance.
Q: Tell us more about the new data that you’re bringing to light on the homeowners insurance market.
A: My coauthor and I have used data on over 70 million mortgage payments to get a real picture of average homeowners insurance premiums by zip code across the United States from 2014–2024. Our data are uniquely detailed, with most existing statistics on homeowner premiums being at the state level and several years out of date.
This new working paper provides an update to the data that we published in 2024. One of the new things we’ve done is make our data publicly available. We expect to be able to update our dataset with 2025 premiums soon.
Q: Homeowners insurance premiums in the United States are skyrocketing. What’s behind this trend?
A: The homeowners insurance industry has been put under stress by multiple inflationary factors all hitting at once.
Starting around the pandemic, we saw rapid inflation in construction costs that directly contributes to higher claims for insurers and ultimately higher premiums for homeowners. Around this time we saw a number of large loss events—wildfires in California, severe hurricanes around the Gulf—that caused insurers to reevaluate the severity of these catastrophic risks and start to reduce their exposure to these markets. Sure enough, we find that construction cost inflation can explain much of the rise in premiums between 2020 and 2024, but that premiums have risen the most rapidly in the highest risk areas.
One of the things that we’ve focused on in our new research is how the cost of insuring catastrophic “cat” risk has changed in particular—think the largest wildfires or hurricanes that affect many policyholders at once. For these perils, insurers often need to buy insurance themselves, either by holding lots of capital, purchasing reinsurance from an even bigger insurer, or by purchasing catastrophe bonds. We’ve seen the cost of this capital surge between 2020 and 2024, and that’s led to higher premiums for homeowners who face cat risks.
Q: What are the impacts of these rising premiums on homeowners?
A: We see two things happening in the cat-exposed markets where premiums are increasing.
First, we’re seeing slower home price growth. In fact, these premium and home price effects are strongest where climate models suggest that risk is increasing. Altogether, this suggests that insurers are repricing risk, and that homeowners expect higher premiums to continue.
Second, we see that insurers are increasingly nonrenewing policies in these same markets. Insurers are adjusting not just by increasing prices but also by reducing their exposure.
Q: Can we mitigate some of the effects we’re seeing now?
A: Ultimately, if climate risk continues to increase and we continue to expand in risky areas, premiums will likely continue to increase. We do have two ideas in mind that could slow this growth.
First, stronger building codes for more durable roofs and wildfire-resistant housing have been shown to reduce risk, which should lead to lower premiums.
Second, I have a new proposal out with Benjamin Collier (fellow risk and insurance faculty at the Wisconsin School of Business) and coauthor Keys about how a federal reinsurance program that we call US Re could sustainably lower premiums in some of these cat-exposed markets. The federal government has lower cost of capital than private reinsurers, which means they could bear the risk of the most severe wildfires and hurricanes more efficiently. However, US Re would need to price its contracts to not subsidize high-risk areas. Ideally, we think that US Re’s coverage should only kick in for the most extreme losses, leaving a large role for private reinsurers. As always with public policy in insurance markets, the devil is in the details.
Tags: