In the six months since the Palisades and Eaton fires burned through Los Angeles, the rebuilding conversation has centered on materials. Fire-rated roofing. Ember-resistant vents. Noncombustible siding. All important, all necessary, and all addressing the second most influential variable in whether a house survives a wildfire.
Not your roof, not your vents, not whether anyone cleared the gutters. Distance. How far away the next house is.
A UC Berkeley research team led by mechanical engineering professor Michael Gollner and postdoctoral researcher Maryam Zamanialaei trained a machine learning model on CAL FIRE's damage inspection database, which catalogs on-the-ground surveys of every structure damaged or destroyed in major California wildfires since 2013. They augmented it with geospatial data on building spacing, construction materials, and vegetation density. The model predicts whether a given structure will survive a wildfire with 82% accuracy, and the variable that matters most, more than roofing material, more than vent type, more than whether anyone cleared the gutters, is structure separation distance.
Densely built areas lose more houses. Not because each individual house is worse, but because fire treats a row of closely spaced structures as a single fuel load. One ignition becomes two becomes ten, and by the time the fire department arrives there isn't a structure to save, there's a block.
The $2,780 Fix Nobody Is Talking About
The Berkeley team found something else that deserves a bigger headline than it got: clearing vegetation within a five-foot perimeter of a house, replacing mulch and bushes with rock or bare soil, reduces structure losses by 17%. On its own, without changing a single building material, without touching the roof, without replacing the windows.
IBHS, the research nonprofit funded by the insurance industry, estimates the cost of establishing an ember-resistant zone within five feet of a home plus enclosing the underside of a deck at roughly $2,780 for a 1,750-square-foot single-story house.
Let me do the math nobody has published. Take a $700,000 home in a California WUI zone with a reasonable 10% probability of significant wildfire exposure over a 30-year mortgage. Expected loss: $70,000. A 17% reduction in that probability is $11,900 in expected value. Investment: $2,780. That is a 4.3-to-1 return before you factor in what it does to your insurance premium, and insurance is where this story gets ugly.
642,000 Californians Have Already Lost the Insurance Lottery
As of September 2025, 642,010 homeowners are insured through the California FAIR Plan, the state's insurer of last resort. Ninety percent of them live in counties that FEMA rates as "Relatively High" or "Very High" for wildfire risk, the same classification bands that private insurers now use as triggers when deciding whether to continue writing homeowner policies in a given area. Private insurers have pulled back from 46 of California's 58 counties, declining renewals and refusing to write new policies in communities where catastrophe models project wildfire losses that exceed what any reasonable premium structure can sustain.
Contra Costa County went from 1,094 FAIR Plan policies in 2021 to 12,837 by September 2025. That is a 1,073% increase in four years. Solano County grew 799%, and Riverside, already among the state's most fire-prone areas, grew 509%. Several of the fastest-growing counties are in the San Francisco Bay Area and Central Valley, regions that were not historically considered wildfire zones.
FAIR Plan coverage is not equivalent to real insurance: it excludes theft, liability, and water damage, the categories of loss that routinely accompany wildfire events and that homeowners most expect their policy to cover. Homeowners who want those coverages must buy a separate difference-in-conditions policy from a private insurer, assuming they can find one willing to write in their ZIP code. Total cost often exceeds what a standard private policy would have charged before the insurer pulled out of the market, making the FAIR Plan not merely inferior coverage but more expensive coverage.
Insurance Commissioner Ricardo Lara has called the path to a healthy market "a 3-to-5-year project." Consumer Watchdog estimates the potential FAIR Plan shortfall could produce surcharges of $975 to $3,700 per California homeowner, not just those on the FAIR Plan. Its reinsurance backstop is $2.5 billion. LA fires alone are expected to generate $30 to $40 billion in insured losses.
This is the context in which $2,780 for a five-foot ember-resistant perimeter stops being a home improvement project and starts being an economic survival strategy.
KB Home Built 64 Houses to Prove the Standard Works
In Escondido, outside San Diego, KB Home completed the country's first "Wildfire Prepared Neighborhood" under the IBHS standard. Sixty-four homes, starting around $1 million for roughly 2,000 square feet. A second project near Sacramento, called Stone Canyon, starts in the high $700,000s.
Where IBHS goes beyond California's building code is specific and measurable. It requires at least 10 feet of separation between structures, and it eliminates what the institute calls "connective fuel pathways," the vegetation, fencing, and stored materials that act as fire bridges between buildings. California's code addresses neither.
Steve Ruffner, KB Home's president for Southern California, changed the Escondido project's design guidelines on the fly after watching an IBHS demonstration in which a home built to standard code burned in about 30 minutes while a home built to IBHS standards did not ignite at all. None of it significantly altered the development's bottom line. In one case they reduced costs: tempered windows required by the IBHS standard broke less often during construction, saving money on replacements.
| Standard | Cost vs. Traditional (1,750 SF) | Key Additions |
|---|---|---|
| CA Code (CWUIC Part 7) | +$13,000 | Ember vents, metal gutters, fire-rated wallboard, tempered windows |
| IBHS WFPH Base | +$9,000 | Same intent, different material choices (saves $4K vs. code) |
| IBHS WFPH Plus | +$15,000 | Enclosed eaves, noncombustible soffits, double-tempered windows |
| 5-ft ember zone + deck | +$2,780 | Rock mulch perimeter, enclosed deck underside |
The HOA agreement in Escondido requires homeowners to maintain vegetation spacing over time. Ruffner said homeowners have been eager to comply, because insurability in California has become existential. If your home is insurable on the private market, you pay what the market charges. If it is not, you go to the FAIR Plan, pay more, get less coverage, and pray the $2.5 billion reinsurance backstop holds.
Around a dozen additional projects are now being designed to the IBHS standard, some with several hundred homes in a single development. IBHS CEO Roy Wright put the logic plainly: "Every time one more structure doesn't burn, that means that structure is not sending off its flame. It's not sending off its embers. Every time we save a structure and it survives, we really narrow the path of how that fire will propagate into a neighborhood."
The Neighbor Problem
The Berkeley model surfaces a truth that individual home hardening cannot fix on its own. You can install a Class A fire-rated concrete tile roof, ember-resistant vents, noncombustible siding, dual-paned tempered glass windows, and fire-rated wallboard on every exterior surface of your house. You can spend $13,000, or $27,000, or $44,000 on materials alone. And if the house eight feet away, the one with a wooden fence connecting to your property and cedar bark mulch piled against its foundation, catches fire, the radiant heat flux through that eight-foot gap can ignite your structure anyway.
Gollner's companion fire spread model, tested against the 2017 Tubbs and Thomas fires, found that more than 30% of structure ignitions were caused by embers, airborne pieces of burning material that can travel hundreds of feet on the wind. Your neighbor's wooden fence becomes your ember source. Your neighbor's untreated deck becomes kindling that launches embers onto your treated roof. Fire does not care whose property line it crosses.
This is why the IBHS standard is a neighborhood standard, not a home standard. A single house built to the strictest specifications surrounded by houses built to none is a marginally safer house in a neighborhood that will still burn. Berkeley's model is measuring a community variable, not an individual one, and the insurance industry is pricing accordingly, penalizing entire ZIP codes rather than rewarding the few homeowners who have individually hardened their properties.
What This Means If You're Building or Buying
If you are building a new home in a California WUI zone, IBHS WFPH Base certification adds roughly $9,000 in materials to a $500,000 construction budget while potentially saving $4,000 compared to code compliance alone, because IBHS achieves comparable fire resistance through different material selections. Add $2,780 for the ember-resistant perimeter. That addresses the single most cost-effective mitigation the Berkeley study identified.
If you are buying an existing home, ask how far the nearest structure is. Berkeley's model makes structure separation distance the single strongest predictor of survival, and that is a fixed attribute of the lot you are purchasing. You can upgrade your roof and your vents after closing. You cannot move your neighbor's house.
If you already own a home in a fire zone, the five-foot perimeter is the intervention with the highest ratio of benefit to cost in the entire Berkeley dataset. Remove wood mulch within five feet of the house. Replace it with rock, gravel, or bare soil. Enclose the underside of any deck. Remove anything combustible stored against exterior walls. That is $2,780 worth of work against a 17% reduction in your probability of losing the structure.
And talk to your neighbors. What Berkeley found is not about your house. It is about the distance between your house and theirs, and about what fills that gap. A neighborhood where every home maintains a five-foot noncombustible perimeter and every fence is metal instead of wood is categorically different from a neighborhood where one homeowner hardens their property and hopes the fire respects the lot line.
Limitations
Berkeley's model was trained on California wildfire data and may not generalize to other fire regimes. Eighty-two percent accuracy is strong for a regional-scale model but means 18% of structures are misclassified, and the model does not account for real-time firefighting response, which can alter outcomes substantially. Structure separation distance being the most influential variable does not mean it is the only variable; construction features and vegetation density contribute measurably, and the model's finding is that all three together produce the greatest reduction in losses.
The 17% figure for five-foot perimeter clearing is a statistical reduction across the full dataset, not a guarantee for any individual property. Topography, wind exposure, and vegetation density all matter, and properties in different conditions may see larger or smaller benefits. IBHS's $2,780 cost estimate applies to a specific home size and site configuration; actual costs vary with lot geometry and existing landscaping.
And the largest limitation of all: structure separation is functionally a zoning and subdivision design variable, not a homeowner variable. For the millions of existing homes in California WUI zones built at five-to-six-foot side setbacks, the Berkeley model's most important finding identifies a problem they cannot individually solve. Only new construction, redevelopment, or coordinated neighborhood-level action can address the gap.
KB Home has shown that building to the higher standard costs roughly the same. Insurance companies are making the economic case inescapable through nonrenewal notices and surcharges that punish density, and machine learning has now identified the variable that matters most, one that no amount of individual home hardening can fully compensate for. What remains is the part that technology cannot do: convincing a state full of homeowners, builders, and zoning boards that the space between houses is worth as much as anything inside them.