My mother kept books for a plumbing supply company in a building that smelled like pipe dope and carbon paper. She taught me that money is always somewhere. It does not disappear between accounts. When a check is written but not yet deposited, the funds exist in transit. Bankers call this "float." Somebody holds it. Somebody pays for holding it. The interval between sending and receiving is a product, and it has a price.
Climate adaptation in the United States has a float problem worth examining in the plainest terms available, which are financial ones.
The federal government identifies a risk. It creates a program. It opens an application window. Then the program moves at the speed of federal procedure while the risk moves at the speed of weather. The difference between those two speeds produces a carrying cost. That cost lands on households, municipalities, school districts, and hospital systems that did not set the timeline and cannot accelerate it. The costs appear in no single federal ledger. But they are in the public record, scattered across program data, census tables, rulemaking dockets, and municipal budgets, if you're willing to do the arithmetic.
What follows is the arithmetic, across four domains where the documents make the gap measurable. The figures are incomplete. Some are estimates. Some don't exist at all, which my mother would have called a finding.
I. The Buyout Queue
When a flood destroys a home in a place where rebuilding is unwise, FEMA's Hazard Mitigation Grant Program can fund a buyout. The local government acquires the property at pre-disaster fair market value. The homeowner walks away. The land becomes permanent open space. Rational outcome. The processing time to reach that outcome, according to NC Division of Public Safety, runs "anywhere from a few months to over two years." A 2019 analysis reported by the Washington Post found a median of about five years from flood disaster to completed buyout. FEMA told the same outlet that almost 80% of acquisitions are approved within two years.
Take the generous number. Two years. During those two years, the mortgaged homeowner continues to owe principal, interest, property tax, and insurance on a home that may be uninhabitable.
After Hurricane Helene hit western North Carolina in September 2024, more than 800 households applied for buyouts. By December 2025, nearly 600 applications had been forwarded to FEMA. None had been approved.
Here is what it costs per month to own a home in those counties, drawn from ACS 2024 five-year estimates for median selected monthly owner costs with a mortgage:
| County | Monthly cost | 12-month total | 24-month total |
|---|---|---|---|
| Buncombe | $1,785 | $21,420 | $42,840 |
| Henderson | $1,608 | $19,296 | $38,592 |
| Yancey | $1,378 | $16,536 | $33,072 |
These figures do not include the cost of living somewhere else. They do not include storage, mold remediation, legal fees, or credit damage. They are the baseline cost of owning the asset the government has agreed, in principle, to take off your hands.
Now set the other clock next to it. Fannie Mae's servicing guide authorizes disaster-related mortgage forbearance in increments of up to three months, capped at 12 months cumulative without special written approval. The Washington Post documented a Helene-affected household whose mortgage company had paused payments for one year. The pause was expiring. The buyout had not been approved. The household still owed $270,000 on a house no longer safe to occupy.
Forbearance clock: 12 months. Buyout clock: two to five years. The gap between those two clocks is the float. It is carried entirely by the household, in the form of resumed mortgage payments on a condemned asset, or in the form of foreclosure, which transfers the cost to the lender and the borrower's credit record simultaneously.
"Paying expenses on homes they cannot live in."
Governor Josh Stein's letter to FEMA used those words. That is a precise description of float, written by a man who probably never used the word.
II. The Heat Standard
On October 27, 2021, OSHA published an Advance Notice of Proposed Rulemaking for a federal heat injury and illness prevention standard. The comment period closed January 2022. A Small Business Advocacy Review panel convened September 2023. The proposed rule was published August 30, 2024. Comments closed January 2025. An informal public hearing ran from June 16 through July 2, 2025. Post-hearing comments closed October 30, 2025.
As of this writing, no final standard has been issued. The rulemaking is four years and eight months old.
During that interval, the agency's own NPRM cites the following BLS baseline for occupational heat exposure: an average of 3,389 workplace heat injuries and illnesses involving days away from work per year, and an average of 34 worker deaths from environmental heat exposure per year. BLS reported 43 heat deaths in 2022 alone. OSHA's NPRM explicitly states that BLS figures are a likely undercount, noting that California workers' compensation data identified three to six times as many annual heat-related cases as BLS reported nationally.
The agency's own conservative numbers, applied to the agency's own procedural timeline:
| Metric | Annual (BLS baseline) | Cumulative, Oct 2021 – July 2026 |
|---|---|---|
| Heat injuries/illnesses, days away from work | 3,389 | ~16,000 |
| Worker deaths, environmental heat | 34 | ~160 |
Those figures come from the document proposing the rule that has not yet become final. The evidence of harm is accumulating inside the procedural timeline of the agency responsible for preventing it. A bookkeeper would note that the receivable is growing while the payment is still being processed.
Beyond the workplace, a report drawing on CDC's HeatRisk Dashboard counted 119,605 heat-related emergency department visits nationally in 2023, with 92% occurring between May and September. (CDC's live Heat & Health Tracker was under maintenance when checked for this piece; the 2023 figure should be verified against CDC's archived data.) A June 2026 GeoHealth study, as reported by the Guardian, estimated current annual heat-related ED visits and hospitalizations across 53 large U.S. metro areas at about 109,000, projecting a rise to 237,000 by 2040. The study estimated associated healthcare costs for those metro areas exceeding $1 billion annually by that date. Whether that figure captures direct medical costs alone or includes broader economic losses such as lost wages was not specified in the available reporting.
No validated national per-case cost figure for heat-related ED visits was located in primary federal sources during the preparation of this piece. The federal government has been working on a heat standard for nearly five years. It can tell you how many workers are hurt. It cannot, or at least does not, publish a clean annual cost figure for what heat illness costs the healthcare system nationally. The float is large enough to measure in bodies but not, apparently, in dollars. My mother would have had something to say about a ledger with quantities but no prices.
III. The Cooling Ledger
Every summer, the federal government spends money helping low-income households survive heat. Every summer, the permanent fix goes unbuilt. The two expenditures sit side by side in the budget like a recurring charge and a capital investment that never quite gets authorized.
LIHEAP, the Low Income Home Energy Assistance Program, was allocated $4.1 billion for FY2024, with Congress continuing that level for FY2025. Reporting based on NEADA analysis, an advocacy organization representing state LIHEAP directors, estimated roughly 12% goes to summer cooling assistance. That implies approximately $492 million annually spent helping people not die from heat in their own homes. The most recent fully validated LIHEAP Report to Congress available from HHS covers FY2022; later figures should be treated as preliminary.
The DOE Weatherization Assistance Program, which permanently fixes the problem, received $366 million in FY2025 and weatherizes approximately 32,000 homes per year. DOE reports that weatherized households save an average of $372 or more annually. The program has served 7.2 million homes since 1976. Fifty years. 7.2 million homes. 32,000 per year at current pace.
No national wait-list or backlog count for WAP was found in DOE's public materials. We know how many homes get done. We do not know how many are waiting, because nobody publishes the number. The absence is a decision about what to count.
The ratio is plain enough. Nearly half a billion in annual emergency cooling assistance. A weatherization program that permanently fixes 32,000 homes a year. The emergency spending recurs. The permanent fix trickles. The gap between them is carried by households in the form of utility bills they can't pay, indoor temperatures that send them to emergency rooms, and repeated summer crises that one program addresses temporarily and the other addresses permanently but at a pace that would take generations to clear the invisible backlog.
At the municipal level, the float takes forms that would be funny if they weren't dangerous. New York City's cooling-center program receives no dedicated funding, according to the city's Independent Budget Office. The nation's largest city runs its heat-emergency response on donated real estate, organizations and businesses volunteering existing air-conditioned spaces during heat waves.
In Philadelphia, 86 schools closed early during a September 2023 heat wave because classrooms lacked adequate air conditioning. The district did not expect adequate cooling in all schools until at least 2027. A $200,000 donation from an NFL quarterback bought AC units for 10 schools and nearly 200 classrooms. As of August 2024, 63 district schools still had inadequate cooling. The capital backlog in a major public school system was being addressed, in part, by the charitable impulse of a man who throws footballs for a living. Call that what you will. It is the receipt you get when no cooling policy exists.
IV. The BRIC Gap
FEMA's Building Resilient Infrastructure and Communities program funds pre-disaster mitigation: flood barriers, wildfire-resistant infrastructure, stormwater systems. The kind of work that costs less before the disaster than after. In April 2025, approximately $3.6 billion in BRIC funds were halted. A federal judge ordered FEMA to make funding available. FEMA reopened BRIC applications on March 25, 2026, with $1 billion available and an application deadline of July 23, 2026. Three weeks from today.
The restored program no longer funds hazard-mitigation planning or non-financial direct technical assistance, which AP reported could disadvantage smaller communities with fewer resources. The communities least equipped to write a federal grant application lost the program component designed to help them write one.
| BRIC timeline | |
|---|---|
| ~$3.6 billion halted | April 2025 |
| Court-ordered restoration | March 25, 2026 |
| Funding available in reopened program | $1 billion |
| Application deadline | July 23, 2026 |
| Hazard-mitigation planning funded | No |
| Direct technical assistance funded | No |
No authoritative average elapsed time from BRIC award to project completion was found in FEMA, GAO, or OIG materials reviewed for this piece. That number either does not exist in published form or is not made accessible. What is documented is the interval: eleven months from cancellation to court-ordered restoration, plus a four-month application window, plus whatever time elapses between award, engineering, permitting, procurement, and construction.
What communities spent or deferred during the eleven-month halt is, in a sense, the central question of this section, and it is the one the public record does not answer cleanly. Municipal budgets do not typically carry a line item labeled "mitigation work deferred because federal grant was canceled." The cost shows up later, in the form of unbuilt flood barriers when the next storm arrives, or in general-fund expenditures reclassified from other priorities, or in projects that simply stop being discussed. I looked for documented examples of specific communities redirecting budgets during the BRIC pause and did not find them in the materials reviewed here. That absence is itself a data point. The float in this domain is real but largely invisible to the accounting systems that would make it legible.
The Sum
Add it up, to the extent the public record allows.
Buyout-pending households in western North Carolina carrying $1,378 to $1,785 per month on properties they cannot inhabit, with mortgage forbearance expiring before buyout approval arrives. Roughly 16,000 occupational heat injuries and 160 worker deaths accumulated during the interval of an unfinished OSHA rulemaking, by the agency's own conservative baseline. Approximately $492 million per year in federal emergency cooling assistance flowing to a problem that a $366-million-per-year weatherization program addresses permanently at a rate of 32,000 homes per year. An eleven-month gap in the federal pre-disaster mitigation program, restored at reduced scope with no published data on how long projects take to complete even when the program is running.
The pattern across all four domains is the same. The federal government identifies the risk. It creates or proposes a program. The program moves at the speed of federal procedure. The risk moves at the speed of physics. The difference between those two speeds is the float, and it is carried by whoever is closest to the danger.
When the carrier can't hold the float, it migrates. The household that can't carry the mortgage on a buyout-pending home faces foreclosure, and the cost transfers to the lender and the borrower's credit history at the same time. A school district that can't fund HVAC replacement closes classrooms, and students and families absorb the loss in instruction time. The county that can't fund interim mitigation takes the next disaster's damage at full price, which was the price the mitigation program was designed to reduce.
None of this is hidden. The timelines are published. The program rules are public. The forbearance limits are in the servicing guide. The rulemaking docket is online. The grant opportunity has a posted closing date three weeks from today. Every component of the float is documented somewhere, by someone, in some system designed for another purpose. What does not exist is a single document that adds them together and names the total for what it is: the cost of the gap between knowing and doing, carried by the people the knowing was supposed to protect. My mother could have balanced this ledger in an afternoon. She would have wanted to know why nobody had.
- Colorado River rules expire: Several operating agreements governing the Colorado River, including the 2007 Interim Guidelines and 2019 Drought Contingency Plans, expire at the end of 2026 with post-2026 guidelines still under development, creating another float between expiring authority and whatever replaces it.
- School buildings as climate infrastructure: A 2020 GAO report estimated that 41% of U.S. school districts needed HVAC updates in at least half their schools, representing roughly 36,000 buildings asked to serve as cooling shelters, air-filtration sites, and classrooms simultaneously on maintenance budgets designed for none of those roles.
- Manufactured housing and split authority: CFPB research found that manufactured-home residents with chattel loans faced higher denial rates, higher interest rates, and lower refinancing likelihood than mortgage borrowers, a financing structure that constrains adaptation for people who own their homes but not the land, drainage, or electrical infrastructure beneath them.
- Cooling center barriers documented: A CDC study of Maricopa and Yuma County older adults found that only 36% knew where a cooling center was located and 18% said electricity cost sometimes or always prevented air-conditioning use, suggesting emergency cooling programs may not reach the populations most at risk of heat illness during the interval before permanent weatherization arrives.