I spent five years on cargo ships and I watched men do everything the procedure said — harness on, clipped in, cargo lashed by the book — and get hurt anyway, because the sea was running a different schedule than the manual. The manual wasn't wrong, exactly. It was written ashore. I've been thinking about those men since I read about Bob Herber.
Herber lives in Heber-Overgaard, a community of about 2,900 year-round residents strung along State Route 260 in Navajo County, Arizona, up in the ponderosa pine at 6,600 feet. He'd been there roughly eight years when, in September 2025, his homeowners insurer mailed him a 30-day nonrenewal notice. Wildfire risk. The policy would run out and the company wasn't going to write another one.
So Herber replaced his roof. He cleared brush and created defensible space around the property, which is the mitigation work every brochure and county plan and fire district in the West tells you to do. He did it before or during his hunt for a new policy, not years later as an afterthought. Another Heber-Overgaard resident, Robert Borucki, told the same FOX 10 Phoenix reporter that his roof alone ran about $10,000. Herber's roof cost wasn't reported, and neither was the brush clearing. Defensible space means the whole perimeter of your place, the immediate zone out to thirty feet and the reduced-fuel zone out to a hundred, and it isn't a one-time job. You maintain it or it grows back. That's money and labor a man puts into the ground on the understanding, stated or implied, that doing the right thing gets noticed by the people who decide whether he keeps his house insured.
Then he started calling around. About half the companies wouldn't quote him at all. The other half wanted premiums five to ten times what he'd been paying. The FOX 10 report that aired in September 2026 is the only public account of his situation, and it didn't describe the terms of those quotes. That matters more than it sounds like it should. The premium is the number everybody argues about, but the dwelling limit, the deductible, the exclusions, the cap on debris removal, the loss-of-use provisions — those are what decide whether a policy does anything for you when your house is burning. We don't know what Herber was being offered for five to ten times the money. We don't know whether it was a working policy or an expensive piece of paper.
We also don't know whether he got coverage in the end, went bare, or took something stripped down. We don't know if he's carrying a mortgage that requires insurance, or what his lender would have to say about any of it. We don't know his age, whether he's retired, whether anybody else lives in the house. The FOX report didn't say and nothing else under his name has surfaced. I'm listing the holes because they're doing as much work here as the facts are. The complete version of Bob Herber's situation exists in exactly one place, which is his own head, and no institution has ever asked him to write it down.
What He Moved Into
Heber-Overgaard has been living with fire a lot longer than Herber has. The Rodeo-Chediski Fire in 2002 burned nearly 469,000 acres across Navajo County, Apache County and the Fort Apache Reservation, took out more than 450 houses and pushed over 2,000 people out. That fire is the founding event of the place as it now exists. Everything since gets measured against it.
When Herber moved in around 2018 or 2019, he was moving into a community that had spent the better part of two decades rebuilding its relationship with fire. The Heber-Overgaard Fire District runs twelve full-time firefighters, seven of them paramedics, with structural engines, wildland engines, water tenders, brush trucks and ambulances, plus mutual-aid agreements with three neighboring districts. In 2019 the Insurance Services Office rated it Class 3 on a scale where 1 is best and 10 means essentially unprotected. Top tier, nationally. The district's own website points out that insurers may use that rating when they price coverage.
There's water, too. Arizona Water Company's Overgaard system runs five active wells serving over 4,500 residential connections. Its 2024 regulatory filing lists fire-flow requirements of 500 to 4,000 gallons per minute for two to four hours, depending on whose standards the local fire authority applies. Whether that flow shows up at Herber's nearest hydrant during a structure fire and a wildland fire at the same time is a question the filing doesn't take up, because filings describe the system on paper. The system under duress is a separate matter and nobody publishes that one.
Herber bought into professional firefighters, a top-tier protection rating, a water system with documented fire-flow capacity, twenty years of institutional memory about wildfire, and neighbors who took the risk seriously enough to thin and clear and plan. He put in his own share — the roof, the brush, the maintenance. For some number of years the arrangement held. He had insurance. The premium was whatever it was. The place worked.
Then the insurance market looked at all of it and said no. The fire district hadn't gotten worse; the water system hadn't failed; Herber had just finished paying for a new roof. Every local system was holding. The one that pulled back was the distant one, run by actuaries and underwriters working from models in offices a long way from the ponderosa pine, and what it sent him was a 30-day letter. Whether anybody from the company ever walked his property line, measured his hydrant distance, or looked at that roof before the decision got made, the public record doesn't say.
Sixty Percent Empty
A single fact about Heber-Overgaard bends every other question in the story. In the 2000 Census, 59 percent of the community's housing units were classified as seasonal, recreational or occasional use. By the 2020–2024 American Community Survey, an estimated 59 percent were vacant. Different category, not a clean comparison. But the picture hasn't moved in twenty-five years.
Six houses out of ten sit empty most of the year. That was the architecture of the place when Herber arrived, and it was the architecture when Rodeo-Chediski burned. It predates the insurance trouble by decades.
The street goes quiet in October and doesn't really come back until Memorial Day. The neighbor who might help you clear brush or notice smoke coming off your eaves is in Phoenix or Tucson or California. Through fire season, when the risk peaks and the burn line matters most, you're living in a town built for three times the population actually in it. The school district enrolls somewhere between 456 and 490 students, drawn from the year-round families rather than the housing stock. The tax base that pays for roads and services runs partly on seasonal money. And a big share of the buyers in the property market are shopping for a cabin, a weekend place, a someday-retirement. A $15,000 or $20,000 insurance premium means something different to them than it means to a man whose whole life is in the house.
Market data as of August 2026 puts the median sale price around $345,000, slightly above a year earlier. Houses sit about 75 days and close at roughly 95 to 96 percent of asking. Realtor.com calls it a buyer's market. Stable enough in the aggregate. What the aggregate can't tell you is whether an uninsured house sells differently from an insured one, how many buyers walked after getting their own quotes, or what sellers are giving up at the table to close. Nobody publishes a separate median for houses no carrier will touch.
We don't know what Herber paid for his place or what it's assessed at now. His address and parcel number have never turned up in any public account, so the county assessor's records can't be matched to his name with any confidence. The person who knows the full cost of this — the roof, the brush, the old premium, the new quotes, the equity gained or lost, the retirement plan that depends on the answer — is Herber. The institutions making decisions about his community are working with less information than he has.
The Burn Line
The fire district keeps a recorded phone line, updated around 7 a.m. and 7 p.m., that sets the day's burning restrictions off wind, weather, available resources, fire behavior and fuel moisture. That's the rhythm of life up there, a twice-daily reading of conditions, the way a ship's crew checks the barometer. You call it in the morning before you plan the day and again at night to know what kind of night you're in for. Over eight years that's close to six thousand readings. Routine, until you try explaining it to somebody who lives somewhere else.
In June 2026 the Flat Fire burned about 100 acres near Black Canyon after a lightning strike, and Heber-Overgaard residents were put in "READY" status, the first tier of the three-tier evacuation system. Be aware, be packed, don't go yet. In 2021 the Wyrick Fire burned 7,592 acres twelve miles northwest and put parts of the community under full evacuation orders. No houses lost either time.
The county's own hazard mitigation plan rates every wildland-urban-interface area in Navajo County at high wildfire risk. The same plan credits thinning and prescribed burning with helping head off a major long-duration fire in recent years. Both of those are true at once, which is the part that never fits in a headline: the risk is high and the work has bought time. The plan doesn't claim the risk is solved. In spring 2026 the county and the fire district ran a bottled-water drive for firefighters, emergency crews and residents ahead of the season. From inside, that's just something you do in May. From outside, it's a town organizing its civic calendar around a threat that never entirely leaves.
What the State Is Trying to Count
Arizona's Department of Insurance and Financial Institutions has issued a data call requiring insurers to report policy-level information on residential properties in state-designated wildfire-risk areas: what they issued, renewed, canceled and nonrenewed, plus premiums, deductibles, coverage limits, claims and the stated reasons for adverse actions. Individual company submissions stay confidential. The state has to publish aggregated, anonymized findings by December 31 each year. The first report under the current call hadn't come out as of this writing.
DIFI's January 2026 task-force materials show 28,630 modeled buildings within designated wildfire areas in Navajo County. That confirms the county is included. It doesn't confirm that Herber's property falls within the mapped boundaries.
A peer-reviewed University of Arizona Extension analysis found strong correlations between wildfire activity and nonrenewals in some Arizona counties, while warning that the available data can't identify why any individual policy was dropped and can't establish causation. I'd rather say that plainly than let it slide by. A resident's mitigation work, an insurer's internal property score, a company's decision to shed exposure across a whole region, and whatever reason got printed on the notice are four different pieces of evidence. Collapsing them into "he did the work and they dropped him anyway" feels right and isn't something the data will support. What the data supports is that nonrenewals are climbing in fire-prone parts of Arizona and that cases like Herber's stay unexplained at the level of underwriting detail.
When the aggregate report lands, it will show trends across designated areas. It won't name his carrier, verify his quotes, or account for why a man with a new roof, cleared brush and a Class 3 fire district got a letter.
The Kitchen Table
I've spent a lot of years watching institutions document their own piece of a problem and walk away pleased with themselves. The fire district documents its readiness, the water company its wells, the Census the vacancy rate. The state is trying, to its credit, to capture the insurance withdrawals. The county's hazard plan covers the fires and the thinning. Each of them does the job. Each job stops at the edge of its own jurisdiction, and the ground between the edges belongs to nobody.
The one place all the pieces meet — fire risk, nonrenewal, property value, a town that's mostly empty nine months of the year, hydrant pressure, the twice-daily burn line — is a kitchen table in Heber-Overgaard. The man at that table has to add up columns nobody will put on the same page for him. Some of the numbers they won't hand over. Some of the projections none of them will make. The state's report is months out. The market data can't tell his house from the seasonal cabins. The fire district can give him the ISO rating and can't promise it'll stop the next carrier from dropping him. The county can tell him the risk is high and the mitigation is working, and leave him to figure out what that means for a retirement, a mortgage, and whether this is still home.
He replaced his roof, cleared his brush, did what the manual said. The manual was written ashore.
- Arizona's first wildfire insurance report: The state's Department of Insurance must publish aggregated insurer data on nonrenewals, premiums, and coverage in wildfire-risk areas by December 31, 2026, the first numbers that could show whether Herber's experience is an outlier or a pattern across Navajo County's 28,630 modeled buildings.
- Mitigation credit versus mitigation recognition: A University of Arizona Extension analysis found strong correlations between wildfire activity and nonrenewals but warned the data cannot show whether individual mitigation work — a new roof, defensible space, Firewise participation — actually changes an insurer's decision on any particular property.
- California's FAIR Plan rate increase: On October 15, California's insurer of last resort will raise dwelling rates by an average of 29.1 percent, with some wildfire premiums potentially doubling — a parallel test of what happens when the backup market reprices risk in fire country.
- Ranchers financing the gap themselves: In western Colorado, Scott and Monte Snyder spent nine hours and eight truck trips hauling water for 800 cattle during the snow drought — another household absorbing the cost of a systemic withdrawal in daily labor that no reimbursement program fully covers.