Note: This piece originally ran on our Substack at ARC. Make sure to sign up for that as well to keep up with all things Earth systems stabilization and the work we do scouting, seeding, building, and scaling stabilization-focused programs. The post was written by Max Berger and Oliver Hollingsaeter; I (Nick) provided editorial notes and guidance.

A UH-1 Iroquois helicopter hovers over destruction caused by Hurricane Andrew. Courtesy the National Archives (Office of the Secretary of Defense).
The 2025 Atlantic season produced three Category 5 hurricanes, including Melissa, which decimated Jamaica, leaving a trail of physical damage equal to about 41% of the country’s GDP in its wake. Yet no hurricane made landfall in the continental U.S. that year, the first such hurricane season since 2015. And 2026’s first hurricane didn’t form until October, marking the latest start to ‘hurricane season’ in 121 years. For a coast still rebuilding from Helene and Milton, it felt, for most of the year, like a reprieve. That is, until Hurricane Isaias struck the Florida Panhandle as a Category 2 on October 9.
But the short moment of calm before Isaias, however welcome, doesn’t mean all is well in regions that are regularly hard hit by hurricanes and have been decimated by them in the past. The most comprehensive study of U.S. hurricane mortality to date estimates that the average storm causes…
…between 7,000 and 11,000 excess deaths over the fifteen years that follow it. Compare that to the average ‘official’ death toll of 24, which only includes deaths NOAA attributes directly to a storm (i.e., deaths that occur during the storm itself, not in following days). If the 7,000-11,000 estimate is even in the right ballpark, the storms of 2017, 2022, and 2024 are still claiming lives along the coast today, and will continue to for years, however quiet the seasons in between.
This introduces several incongruencies and challenges, not just to conventional thinking about hurricanes, but (and likely more importantly) also to how to proactively plan, prepare for, and potentially prevent damage and loss from these storms. Hurricanes are typically treated as acute challenges, situating them in the realm of emergency managers, disaster preparedness planners, and insurers. They are that. And they also fit squarely into many other practitioners’ remits, including practitioners with much longer time horizons, even if that isn’t currently standard operating procedure. Put simply, hurricanes are broader social policy issues, and therefore warrant investments in mitigation, even if, in practice, that predominantly looks like small-scale R&D today. In the future, those investments could reasonably yield large-scale efforts to better predict, prepare for, and even suppress catastrophic storms before they make landfall. If successful, this would deliver massive ROI on early catalytic capital allocation today.
This discussion drives four main observations we’ll explore:
The bill: Hurricanes are the costliest disaster type in the U.S. record, with costs clocking in at roughly $1.5 trillion since 1980. By the best recent estimates, they’re also a leading and largely uncounted driver of mortality along the coast.
The pace: Warming is changing storms faster than scientists can understand and institutions can adapt to them. Investment in readiness and resilience isn’t keeping pace either. Florida’s FY 2024-25 budget included ~$550 million in identifiable state-administered resilience and hazard-mitigation appropriations (federal funding included) whereas three hurricanes that year produced ~$8.33 billion in reported insured losses, yielding a loss-to-investment ratio of ~15 to 1. Recent years tell an even more distorted story: Hurricane Milton (2024) alone caused an estimated $34.3 billion in economic damage.
The burden: The costs of hurricanes often hit those least able to bear them hardest, particularly through health, housing, credit, and public services vectors.
The lever: Reducing the hazard hurricanes themselves present may be the most progressive intervention available, insofar as a weaker storm protects everyone in its path. Yet, translational R&D focused on potential interventions that could disrupt hurricanes before they make landfall is critically underfunded.
Beyond property damage

Winnie, Texas, September 18, 2008 — A displaced resident sitting in a Red Cross shelter eating dinner by himself. Hurricane Ike ravaged towns along the Gulf of Mexico and left a broad path of destruction in its wake. Courtesy the National Archives (Mike Moore/FEMA)
Most accounting of hurricanes starts and stops with dollar figures, which, while large enough in their own right, are far from comprehensive in accounting for the ‘true’ costs of hurricanes on society. Between 1980 and 2024, NOAA counted 67 U.S. tropical cyclones that each caused at least $1 billion in damage; together they cost $1.54 trillion, at an average of $23 billion per storm (adjusted for inflation), more than any other disaster type in the U.S. Worldwide, cyclones caused about $135 billion in losses in 2024 alone, representing more than 40% of all natural catastrophe losses that year. Here’s one initial example of clear and tangible damages that these figures don’t even capture, however: National incomes stay below their prior trend for two decades after a major storm. For smaller economies, a single storm can be existential (Hurricane Maria’s damage and losses in Dominica in 2017 equaled 226% of the island’s GDP).
Beyond the headline dollar figures, another cost that is systemically left out is the human toll. Rachel Young and Solomon Hsiang’s analysis of 501 storms found that hurricanes elevated mortality for fifteen years after any given storm. The effect is so pronounced that the ‘hurricane climate,’ i.e., how often and how severely storms tend to strike a given place, accounts for an estimated 3.2–5.1% of all deaths in the contiguous U.S., and 13% of deaths in Florida. Over the period they studied, that amounts to more deaths than every U.S. motor vehicle accident combined, as well as all deaths by infectious disease or American battle deaths in war over the same time period.
This excess mortality burden falls heaviest on infants, young adults, and Black Americans, who face roughly three times the hurricane-driven excess mortality risk of white Americans. As the authors of the study put it, the hurricane climate, “…previously thought to be unimportant for broader public health outcomes, is a meaningful underlying driver for the distribution of mortality risk.” (Young and Hsiang, Nature, 2024.)

Young and Hsiang, Nature, 2024.
It’s one study, to be sure, and exactly how a storm translates into deaths years later isn’t yet well understood (a subject for deeper investigation). Still, plenty of additional evidence points towards similar conclusions, and certainly, at a minimum, underscores the importance of considering costs beyond damaged infrastructure and property. After Hurricane Maria, Puerto Rico’s estimated 4,645 excess deaths dwarfed the official death count of 64, and a meta-analysis of 71 studies finds that cyclone exposure raises all-cause mortality (the total number of deaths from all causes within a population during a period of time) by 9%.
Despite the literal and figurative burden of all this evidence, the tools institutions use to set health priorities fail to account for nearly all of what we’ve discussed here, the same way that economic accounting and many other studies, let alone official death figures, do. The Global Burden of Disease study, the standard source for disability-adjusted life years, counts disasters by their immediate deaths and injuries; it counts the 24, not the 7,000. Even if further work were to revise these estimates substantially, the gap between the two numbers is wide enough that any serious accounting of what hurricanes cost ought to reckon with it.
What's changing, and how fast?
Global warming isn’t clearly correlated with an increase in the number of hurricanes globally. However, it’s changing how strong they get, how quickly, and the size and range of areas in which they wreak havoc. For one, rapid intensification (i.e., an increase in sustained winds of at least 35 mph within 24 hours observed in a specific storm) is increasingly observed closer to shore. Such events near the world’s coastlines tripled between 1980 and 2020. Hurricane Milton intensified from a tropical depression to a Category 5 in less than 49 hours, faster than any previously recorded Atlantic storm.
Moreover, warmer oceans appear to be loading storms with more moisture, which keeps feeding them after they come ashore: in the late 1960s, a typical North Atlantic hurricane lost about three-quarters of its intensity in its first day over land, whereas today it loses about half. Hurricanes therefore stay dangerous farther inland, which helps explain how Helene killed around 250 people, many of them in the mountains of western North Carolina, hundreds of miles from the Gulf. Hurricanes have killed far from the coast before, but it’s still striking that a storm making landfall in Florida’s Big Bend, nearly 400 miles from Asheville, would kill more people in North Carolina than in any other state, on its way to becoming the deadliest hurricane on the U.S. mainland since Katrina.
Skeptics will point out, reasonably (and helpfully, as it allows us to broaden the range of our exposition and exploration), that several studies find past U.S. losses show no clear trend once they’re adjusted for growth in population, wealth, and inflation. Others, using different methods, do find one, but even taken at face value, the no-trend finding is less reassuring than it sounds, since it means exposure itself is a large part of the problem. A 2025 reassessment of more than a century of U.S. hurricane damage concludes exactly this: exposure is compounding rapidly in many areas. Florida’s population alone grew by two-thirds between 1990 and 2020, adding over 8.5 million residents, including in areas with particularly pronounced hurricane-exposure risk. Regardless of how climate change will alter hurricanes, the fact that population, infrastructure, and wealth continue to concentrate in high-risk areas highlights a widening gap between where risk is accumulating and what the institutions charged with managing it are equipped to do.
Institutions of nearly every kind, meanwhile, run on slower clocks than either the storms or the people moving into their path. Florida’s statewide building code offers one success story; by rewriting and updating standards, the state cut windstorm losses by up to 72%, but only for new infrastructure and buildings. In 2021, nearly two decades after its adoption, homes permitted under the code amounted to only 28% of the state’s single-family homes, as the housing stock turns over slowly. In most other areas of proactive institutional readiness and resilience efforts, there’s little in the way of success to speak of of late.
Federal pre-disaster mitigation grants have historically saved about $6 for every $1 spent, predominantly by paying for measures like buying out flood-prone homes, installing storm shutters, and building safe rooms. Yet FEMA terminated its flagship program for them in April 2025 and, for nearly a year, awarded no new subgrants until a court forced its restart. In May 2026, a council appointed to overhaul FEMA went further, recommending that the minimum federal share of post-disaster costs fall from 75% to 50%, leaving states to shoulder far more of the bill. Still, we find, storms are changing on timescales that institutions and certainly physical assets, like the housing stock or even the codes by which it’s expanded, were never designed to match.
The appetite for readiness
As introduced in the previous section, the returns on many readiness measures are well documented. Adopting the latest building codes, for one, is estimated to save $11 for every $1 invested, while improvements in forecasting since 2007 have cut the total cost of a U.S. hurricane by an estimated 23%, or about $2 billion per storm.

SGT. Scott Anderson, Company A, 10th Mountain Division (Light Infantry), hammers shingles on a roof as he repairs damage caused by Hurricane Andrew, which struck the region on August 24th. Courtesy the National Archives (Office of the Secretary of Defense).
Why, then, do such relatively straightforward (and high-leverage) measures remain so critically underfunded? The problem isn’t a lack of evidence so much as a lack of political will. Florida’s 15-to-1 ratio of insured losses to resilience spending is, if anything, a floor: the state’s figures exclude federal flood claims and uninsured losses. Munich Re estimated Milton’s insured losses at about $25 billion, more than four times the roughly $5.6 billion insurers reported to the state for that storm. In December 2024, Congress approved roughly $100 billion in disaster aid, after the fact and largely for recovering from damage already done, rather than hardening homes against future storms. When Florida’s home-hardening grants, which pay $2 for every $1 a homeowner contributes, reopened with $200 million, meanwhile, the money ran out in, well, a matter of days. Among households, then, at least when the state shares the cost, the demand for readiness appears to be there; what’s lacking is the institutional willingness to fund it at anything like the scale of the losses.
Political scientists have a well-documented explanation for this dynamic. Andrew Healy and Neil Malhotra found that voters “reward the incumbent presidential party for delivering disaster relief spending, but not for investing in disaster preparedness spending,” even though, per estimates in the same study, each $1 of preparedness avoids about $15 in future damage. Short memories compound the problem, as flood insurance take-up jumps after a flood and then steadily decays back to baseline. Quiet seasons, perhaps unsurprisingly, are precisely when memory fades, among households and institutions alike, and the next budgets get written.
The human ledger
The excess deaths we began with are only the starkest of the costs that go uncounted, and it's here that hurricanes become a much larger, broader social policy problem, insofar as those costs often concentrate in areas and communities least resourced to absorb them, let alone proactively adapt to them.
The excess deaths we began with are only the starkest of the costs that go uncounted, and it’s here that hurricanes become a much larger, broader social policy problem, insofar as those costs often concentrate in areas and communities least resourced to absorb them, let alone proactively adapt to them.
Health: Respiratory hospitalizations among older adults rise about 14% in the week after a storm, and the damage travels well beyond the storm’s path. When Helene flooded the North Carolina plant that made about 60% of America’s IV fluids, hospitals as far away as Boston rationed supply for months.
Mental health: Psychological harm after a hurricane is common and, counterintuitively, can worsen with time. A year after Katrina, PTSD prevalence among the region’s pre-storm residents had risen from 14.9% to 20.9% (with the increase concentrated outside New Orleans), and suicidal ideation had more than doubled, a result the researchers described as “contrary to results in other disaster studies” and traced largely to unresolved stresses over insurance, housing, and rebuilding. Among low-income mothers in New Orleans, one in six still showed probable PTSD twelve years later.
Housing: Nearly one in four displaced renter households never return home, compared with one in ten homeowners, and in hurricane-hit counties rents climb for two years and evictions follow. Homeowners pay through insurance: disaster risk now passes into premiums at roughly twice the rate it did only a few years earlier, cutting home price growth by more than $40,000 in the most exposed ZIP codes. For most families, in other words, the hurricane bill doesn’t arrive with the storm, but rather every month through escrow.

Union Beach, N.J., Dec. 3, 2012 — Pamela Vazquez, Hurricane Sandy survivor, looks out over the destruction that was once her home and her town, before Hurricane Sandy hit. Courtesy the National Archives (Liz Roll/FEMA)
Inequality: Perhaps most troubling, disasters tend to widen whatever gaps they find. In heavily damaged counties, white households gained an average of $126,000 in wealth after recovery, while Black, Latino, and Asian households lost between $10,000 and $29,000. Debt in collections keeps rising for years after a disaster, with the steepest credit damage among those already struggling, and children in affected areas lose learning that doesn’t come back, at a cost in lifetime earnings roughly as large as the property damage itself.
From landfall to balance sheet
Hurricane risk also travels, increasingly, throughout the financial system, whether in the private sector or in public budgets, at all levels of government from counties to countries. Much of it was never insured to begin with: Munich Re estimated that less than 30% of Helene’s losses were covered, and in the inland counties the storm struck, 0.8% of homes had flood insurance.
What is insured has grown fragile. In Florida, as established insurers retreated from high-risk areas, thinly capitalized newcomers filled the gap, and nearly one in five of them later became insolvent, many having carried ratings just high enough to pass their risk into the federally backed mortgage system.
Public budgets absorb much of the rest, often for a decade (and likely even longer). Local tax revenues remain 7.2% below their expected path six to ten years after a major hurricane. And when Tatyana Deryugina counted the rise in unemployment insurance, public medical payments, and other social programs in the decade after a storm, a hurricane’s fiscal cost jumped from about $150 to at least $780 per resident. Hurricanes, in other words, are already a social policy expense; they’re simply booked under other line items.
The need to look upstream
None of this is an argument against adaptation, which works and deserves far more funding than it receives. It is, however, an argument that adaptation alone is not keeping pace. Based on the cost and the absence of sufficient institutional willingness to pay and political will to legislate and allocate more capital out of public coffers, it may not be able to. Hence, reducing the intensity of storms themselves before they make landfall, if feasible, is a prerogative that is well worth researching and belongs in the portfolio of consideration with respect to alleviating the comprehensive burden of hurricanes and storms on society. This is true for three reasons in particular.
1) The damage is concentrated in the strongest storms: Historically, major hurricanes have made up 24% of U.S. landfalls but 85% of the damage. Because damage rises roughly with the cube of wind speed, a 10% cut in peak winds could reduce damage by 25% or more. Mortality, too, rises with the wind a population experiences, though differently.
2) Reducing the hazard offers a progressive strategy: Most other strategies to alleviate the burden of hurricanes on society are neutral to regressive, reaching people unevenly. Florida’s home-hardening grants, for instance, exclude renters, condo owners, and mobile home residents; insurance favors those who can afford premiums; and evacuation favors those with cars and somewhere to go. Large public works, from levees to seawalls, do protect whole communities, but only those behind them, and at a cost that can leave those communities waiting for years: Congress authorized Texas’ proposed coastal barrier system in 2022, at an estimated $34 billion and rising, and federal appropriations had come to about $5.5 million as of early this year. A weaker storm, by contrast, protects everyone in its path at once, and before the damage rather than after. It is about as close to a pure public good as resilience gets, which is precisely the kind of good markets underprovide and social policy exists to supply.
3) The cost of researching interventions is relatively trivial: Whether any intervention (i.e., strategies to weaken, suppress, or redirect storms before they make landfall) can work remains an open question, but answering it is inexpensive relative to nearly everything else on the ledger. The foundational work, from modeling and laboratory studies to small, carefully governed field experiments, would cost a small fraction of what a single average billion-dollar hurricane does (about $23 billion), let alone the $34 billion, and rising, estimated for one Texas coastal barrier. Deploying an intervention at scale would be another matter, and what it would cost isn’t yet known; the point, rather, is that the option to find out is cheap, and its value, if the answer turns out to be yes, would be measured against storms that each cost tens of billions.
This isn’t to suggest the path is short or certain. No intervention method (i.e., strategies to weaken, suppress, or redirect storms before they make landfall) has demonstrated in practice that it can reliably and predictably weaken a hurricane in sufficiently controllable fashion. As the first piece in this series laid out, the science, infrastructure, and governance for this type of intervention all remain to be built. The geopolitical questions are significant and quite uncertain, too. A storm weakened off one coast, or a storm with an altered pathway, still risks making landfall elsewhere.
All that said, a decision-analytic study of South Florida found that, if feasible, modification could reduce losses from an intense storm more than hardening buildings would, while hardening provides “fail safe” protection for average storms. Intervention and adaptation are not substitutes, but complements: as warming strengthens storms, interventions that could reliably reduce their intensity might allow existing and planned defenses to protect against more extreme events than they were designed for. Those with the most to gain, meanwhile, include communities in Jamaica, Dominica, the Philippines, and Bangladesh, where the capacity to adapt is lowest.
Who should act, and how
Governments should count the full bill, i.e., excess mortality, displacement, and a decade of social insurance spending, in disaster accounting and cost-benefit analysis. With that bill in clearer and more consistent view and consideration, they should right-size readiness funding to the losses it is meant to offset, and then grow it in proportion to the rising cost of climate impacts, rather than leaving it to rise and fall with the salience cycle. That investment should also run on two tracks at once: location-specific mitigation and adaptation, from stronger codes to retrofits suited to each region’s exposure to improved drainage infrastructure, and responsible translational R&D on hurricane modification, which, if it proves feasible, would be the highest-leverage option available, requiring significantly less capital while directly reducing risk at its source. Since that science will take years to mature, the slow work of governance must also begin in parallel.
Philanthropists and impact investors are among the few actors able to move first, especially on catalytic funding for novel science and translational R&D toward interventions. These are the actors who are best positioned to take the earliest and riskiest bets in a field before its results can justify larger commitments. Intervention research is likely too early for the fund return dynamics and timelines of venture capital and, as the history of storm modification illustrates, too politically exposed for most agencies to sustain on their own. Philanthropic capital, by contrast, can tolerate early failures and stay through political cycles, funding high-risk, high-reward groundwork, from modeling and small-scale experiments to public engagement, that turns an open scientific question into one that governments and insurers can back. There’s plenty of precedent for this: for example, the Rockefeller Foundation’s wheat research in 1940s Mexico came some two decades before the governments of India and Pakistan, with backing from USAID, imported its high-yielding wheat at scale and helped launch the Green Revolution.
Impact investors, for their part, can follow close behind, backing the observing, forecasting, and engineering tools that have markets of their own today and grow more valuable as the field matures. It’s this capacity to derisk a field for those who follow that makes philanthropy the natural backer of a new, independent entity whose goal is to build, within five to ten years, a scientifically robust, societally legitimate, and operationally ready portfolio of technologies and governance frameworks to weaken hurricanes before landfall, starting with the southern U.S.
Insurers and reinsurers have perhaps the most direct financial stake in weaker storms, and more levers than they tend to use. For one, the tail of the risk distribution is where reinsurers’ exposure sits: because their contracts typically pay out only above a high threshold, even a modest reduction in a storm’s intensity would cut their losses by proportionally more than it cuts total damage. Moreover, the industry holds two assets no one else does: the catastrophe models that set premiums, reinsurance rates, and catastrophe bond prices, and the claims data that could pin down how steeply damage actually rises with wind speed, which largely determines what weakening a storm is worth. It has also shown what it can do with them. The Insurance Institute for Business & Home Safety, funded by insurers, built a facility that tests full-scale homes in hurricane-force winds, and after Hurricane Sally, Alabama homes built or retrofitted to its FORTIFIED standard had as many as 74% fewer claims than those without it.
Perhaps most critically, insurers are the actors best placed to turn verified risk reduction into something that can be priced and paid for. Their contracts are already written in the language of storm intensity: Jamaica’s $150 million catastrophe bond paid out in full after Melissa because a third-party modeler found the storm’s central pressure and path met its triggers, whereas Beryl, which caused widespread damage the year before, didn’t trigger it at all. The same measurement and modeling infrastructure could, in time, be used to estimate the losses a weakened storm avoided, which would open a path for hazard reduction to be financed by those it protects rather than by philanthropy alone. The industry’s expertise in allocating liability, meanwhile, bears directly on the question raised earlier of who pays when a storm weakened off one coast still strikes another, which any modification effort will need to answer convincingly to earn public trust. For insurers and institutional holders alike, co-funding foundational research now would be a modest hedge against exposures of this size, and one that gives the industry a hand in shaping the standards by which hazard reduction is eventually judged.
The calm between storms
Social policy is, fundamentally, how a society decides which risks it will carry together. For most of the last century, hurricanes have been viewed as entirely unpreventable acts of God, the damages and perils of which then become the responsibility of emergency managers and local government (and those impacted, regardless of how underresourced they often are) to deal with and navigate. The evidence now suggests they’re a slow-moving public health, housing, and fiscal crisis, one whose costs continuously arrive for years after the storm that set them in motion.
The relative calm the U.S. coasts enjoyed between Milton and Isaias was never going to last, and it’s precisely in such periods, when memories fade and the next budgets get written, that institutions have the most room to act, whether by counting the full bill, funding readiness in proportion to it, or beginning the research and governance work that might one day make the storms themselves less destructive. The question now is whether institutions will actually use that room, when they have it, to invest in mitigation, adaptation, and intervention, or whether they’ll continue to hope that the future storms they know are coming won’t make landfall for another month or year.
Advanced Research for Climate Emergencies (ARC) is a nonprofit venture studio for Earth systems stabilization. A fund of Renaissance Philanthropy, ARC originates, incubates, and scales translational research and development programs to predict, prevent, and prepare for catastrophic climate risks. Get in touch.


