Lecture 8

Wildfire Risk, Smoke, and Insurance

Byeong-Hak Choe

bchoe@geneseo.edu

SUNY Geneseo

🔥 Forty-One Years of U.S. Wildfire

Each stripe is one year. Darker means further above that row’s 1985–94 average.

What got worse, and what did not?

Data: National Interagency Fire Center.

Open with no definitions. Show the acres row and ask what students see. Click to add the fires row, then ask the question.

Each row is scaled to its own 1985–94 average, so both rows share one colour scale. The series starts in 1985 because the 1983–84 counts look like a reporting break.

Class 1 (50 minutes): trends 10, wildfire risk 9, smoke 28, exit ticket 3.

📈 Not More Fires: Bigger Fires

Since 1985–94: 14% fewer fires, 2.5× the acres, 2.9× the acres per fire.

How can fewer fires burn more land?

Data: NIFC. Lines are decade averages. Acres per fire is total acres ÷ total fires.

The count includes every reported fire, even very small ones. Satellite records of large fires tell the same story: larger and more frequent big fires since 2000.

Single years are noisy. From 2024 to 2025, fires rose 20% while acres fell 43%. Use decade averages for trends.

🏚️ Acres Are Not Damage

• U.S. acres burned in 2025 were 32% below the prior ten-year average.
• Yet January 2025 brought California’s 2nd and 3rd most destructive fires.

What turns a fire into a disaster?

Data: CAL FIRE top 20 list, April 3, 2025 (Eaton and Palisades counts not final); NIFC.

Eaton and Palisades together destroyed about 16,000 structures, with insured losses near $40 billion, yet they burned about 37,700 acres, under 1% of U.S. acres that year. Seven of California’s top 10 fires by structures destroyed came in 2015 or later; the top 10 also includes Tunnel (1991), Cedar (2003), and Witch (2007). “Structures” includes homes, outbuildings, and commercial buildings.

🎯 Wildfire Risk

Companion reading: The Economics of Wildfire Risk and Smoke

Plan for three classes. Class 1: trends, risk, and smoke. Class 2: deciding under risk, then the insurance choice. Class 3: insurance markets, behavior, and recovery. The companion reading follows the same order and has the sources and longer derivations.

🎯 What Is Wildfire Risk?

• Hazard: how likely and how intense a fire is
• Exposure: who and what stands in its path
• Vulnerability: how much harm it does on arrival

\[ \textbf{Wildfire risk} \;\propto\; \textbf{Hazard} \times \textbf{Exposure} \times \textbf{Vulnerability} \]

The product is shorthand. If any one element is zero, there is no loss: a fire in empty wilderness can be a large hazard with no exposure. Hazard has grown with hotter, drier fuels and faster fires; the next two slides show exposure and vulnerability. Smoke later in the class stretches “exposure” far beyond the fire.

🏘️ Exposure: Building in the Bull’s-Eye

• 97% of new interface area came from new housing, not new vegetation.
• Beauty and risk come bundled.

Why do people keep building where it burns?

The wildland-urban interface is where houses meet or mix with undeveloped vegetation. The “bull’s-eye effect”: a growing stock of homes and roads sits in risky, attractive places, so the same fire now reaches more. People buy a bundle of views, schools, weather, and jobs, and the risk comes attached. Connect to hedonic pricing in Lecture 7.

🧱 Vulnerability: Same Fire, Different Damage

• Each extra 10 feet between homes raises survival odds 7–13%.
• Vent screens cost $100–$500.
• Adoption stays low. Why?

Data: IBHS inspections after the January 2025 Los Angeles fires.

The features are noncombustible siding, Class A roofing, and 1/8-inch mesh vents. Clearing fuel near a home costs $500–$2,000. Owners also weigh costs that are not in dollars: hardening can look worse, and clearing removes vegetation they like. These are comparisons, not causal estimates. Class 2 turns to causal evidence on building codes.

💨 Smoke Without Borders

Smoke stretches exposure from a burn perimeter to half a continent. This section uses the 2023 Canadian fire season, which sent dense smoke over New York and the Great Lakes. About 28 minutes, ending with the exit ticket.

🍁 Canada’s Record Fire Season

• 2023: 17.6 million hectares, over 2× any year since 1990.
• About 7× the 1990–2022 average.
• Smoke drifted south and east for weeks.

Whose problem is Canada’s fire season?

Data: National Forestry Database (provincial and territorial fire agencies).

Totals add up what each province and territory reports. Agency perimeters include unburned islands, lakes, and bare ground, so satellite mapping gives 15.0 million hectares, still more than twice the 1989 record measured the same way. Before 2023, the largest year in this series, which starts in 1990, was 7.5 million hectares in 1995; the national record was about 7.6 million in 1989. The 1990–2022 average was 2.4 million. 2025, at 8.4 million, was the second-largest year. A record 232,000 people from 200 communities were evacuated. In Quebec, the provincial agency fights fires across the populated south and mainly records fires north of the 51st parallel unless they threaten communities or key infrastructure.

Most of the burned area was remote boreal forest, so property losses were small relative to the area. Answer to the question: Canada’s, and also everyone downwind. That is where the economics starts.

🌆 June 2023: Smoke Reaches the East

June 7: New York’s average monitor read 153 µg/m³, over 4× the daily standard. One hour topped 400.

Who paid for this smoke? Did anyone send a bill?

Data: EPA Air Quality System. Each line averages a metro’s monitors.

Ask who remembers the orange sky in June 2023. Two waves: June 6–8 hit the East Coast and upstate New York, and June 27–29 hit Chicago and Detroit before moving east. A typical 2023 day in these metros was 6–10 µg/m³. The hourly peak in New York was 412 µg/m³. Rochester’s main monitor stopped reporting complete days during the June 6–7 peak, so Syracuse stands in for upstate New York here; Rochester appears on the next slide.

Answer to the question: the people who breathed it paid, in health, lost work, cancelled plans, and money spent on defense. No one sent a bill, which is the definition of an externality.

📉 Smoke Erodes Clean-Air Gains

• Clean-air rules cut unhealthy PM_2.5 days to near zero.
• Smoke brought them back in 2023 and 2025.
• Nationally, smoke has undone about a quarter of earlier PM_2.5 gains in most states.

The Clean Air Act regulates smokestacks. Whom do you regulate for smoke?

Data: EPA AirData. A day counts when PM_2.5 sets the metro’s AQI above 100. July 4–5 (fireworks) left out.

The early-2000s days came from power plants, vehicles, and industry. Regulation of those sources worked, so most recent years have no unhealthy PM2.5 days. In 2023 and 2025, the unhealthy days came in summer smoke episodes from Canadian fires.

Answer: there is no smokestack. The emitter is a forest in another country, and EPA rules let states set aside smoke-affected days when judging whether they meet the standard. Smoke sits largely outside the regulatory system, so its damages go unpriced.

🫁 Smoke’s Health Bill

• June 2023 asthma ER visits: NYC +44%, rest of New York +82% at the peak.
• To add deaths in dollars, use the value of a statistical life (VSL): about $14 million.

\[ \text{Damage} = \text{Deaths} \times \text{VSL} \approx 4{,}100 \times \$14.2\text{M} \approx \$58\text{B} \]

Both bars come from one study of the 2023 Canadian fires, and they use different methods, so do not add them. The short-term estimate counts extra deaths on smoke days. The long-term estimate applies the year’s added exposure, about 1.5 µg/m³ on the U.S. annual average, to long-run mortality risk. The gap shows how much the answer depends on the exposure-response relationship.

ER visits: in New York City, asthma visits rose from about 182 to 261 a day during June 6–8. In the rest of the state, visits on June 7 were 82% above the June 1–5 average and nearly tripled for ages 10–29. Nationally, asthma visits ran 17% above expected on the 19 smoke days of April–August 2023.

VSL: the federal transportation value for 2025 is $14.2 million; EPA’s recommended value is somewhat lower in the same dollars. With the long-term estimate, the damage approaches half a trillion dollars before discounting. For scale, all wildfire smoke causes about 41,000 U.S. deaths a year (2011–2020), projected to reach about 71,000 a year by 2050 under high warming.

Ask: is a statistical life the same as a life? No: it summarizes willingness to pay for small risk reductions spread over many people.

💼 Smoke and Paychecks

0.1% less quarterly pay for each extra day of smoke

$125 billion of U.S. labor income lost to smoke each year, 2007–2019

• Most of the loss hits people who keep their jobs.
• Nearly 2% of U.S. labor income, before 2023.
• None of it shows up in a fire’s damage tally.

Why might lost pay be larger than the medical bills?

Each additional smoke day cuts quarterly earnings by about 0.1%. Summed over all the smoke Americans breathed, losses averaged about $125 billion a year in 2018 dollars over 2007–2019, nearly 2% of labor income, and that period ends before 2023. Lost jobs and people leaving the labor force explain only 13% of the loss; the rest comes from people who keep their jobs: fewer hours, sick days, caring for sick children, and lower productivity.

Answer: hospital visits are rare, while lost hours touch millions of workers at once. In the study, lost earnings per unit of smoke exceed the mortality cost the authors compute, though that ranking depends on which mortality estimate and VSL one uses.

😷 Defending Against Smoke

• Purifiers, masks, staying in: defensive spending.
• Willingness to pay to avoid smoke ≥ what people spend to defend.
• Who shows $0 in the data?

Classroom numbers.

On smoke days, online searches for air filters, purifiers, smoke masks, and air-quality monitors rise. On days with smoke PM2.5 above 50 µg/m³, the share of people who stay home all day rises by only about 3 percentage points. Protective searches show no detectable rise in roughly the poorest third of counties, while wealthier counties stay home more.

The bound: if you spend $250 to get through a smoke episode, avoiding it is worth at least $250 to you, and the purifier does not remove all the harm. This is the averting-behavior method from Lecture 7. Answer to “Who shows $0?”: renters without central air, outdoor workers, and households that cannot spare the money. A study that counted purifier sales would record no damage for them.

🌍 Pollution Without Borders

• Canada weighs Canada’s benefits.
• U.S. benefits are external.
• So Canada chooses \(R_C\), below \(R^*\).

Exit ticket: Name one cost of the 2023 smoke that no market price ever recorded.

Canada counts benefits to Canadians, including Canadian cities downwind, but not benefits to Americans, so smoke reduction stops at \(R_C\), short of the joint optimum \(R^*\). Why no Coasean deal? Millions of affected people, no court that enforces a cross-border right to clean air, and free riding. The externality runs both ways, since smoke from western U.S. fires drifts north. Cooperation already exists for flames: under a reciprocal arrangement dating to 1982, the countries share crews and equipment, and in 2023 the United States sent about 2,450 personnel, more than any other country. The 1991 Canada–U.S. Air Quality Agreement covers acid rain and ozone, not wildfire smoke; a 2024 joint review recommended updating it.

Keep the graph honest: much of Canada’s burned area is remote boreal forest, where suppression is costly, often futile against the largest fires, and not always desirable, since fire renews these forests. So \(R^*\) may be modest, and U.S. adaptation (clean-air rooms, purifier programs, smoke forecasts) and climate mitigation are complements. Exit-ticket answers: lost pay, asthma attacks, deaths, cancelled games and flights, days spent indoors, discomfort.

⚖️ Deciding Under Risk

Class 2 (50 minutes): mitigation and adaptation 3, expected value and CBA 12, risk aversion and uncertainty 9, building codes and spillovers 10, practice 6, insurance choice 10.

🌎 Mitigation and Adaptation

🏭 Climate mitigation

Cut greenhouse gas emissions.

Benefits are global and slow.

🧯 Climate adaptation

Reduce fire losses where people live.

Benefits are local and sooner.

Careful with words. Fire managers call brush clearing and hardening “mitigation.” In climate economics, those are adaptation.

• Some warming is locked in, so we need both.
• Today: how to choose among adaptation projects.

Flag the vocabulary now, because fire agencies and the studies use “mitigation” in the fire sense. In these slides, “hardening” means making a building harder to ignite. A purifier against smoke is adaptation too.

📐 Expected Value

Under risk, we can list outcomes and attach probabilities.

\[ E[NB] = \sum_i P(x_i) \times NB(x_i) \]

• A 2% chance of $100,000 in damage:
• \(0.02 \times \$100{,}000 = \$2{,}000\) a year
• That is an average, not next year’s bill.

In 98 of 100 years the owner loses nothing; in 2 of them, $100,000. The picture is one possible century, and the fire years could fall anywhere. Stress that $2,000 is never the actual loss in any year.

🌲 CBA for a Fuel-Treatment Project

Fire seasons	\(NB\)	\(P\)
Mild	$5M	0.27
Average	$10M	0.49
Severe	$20M	0.23
Burn escapes	−$100M	0.01

\(E[NB] = \$9.85\) million. Would CBA approve the project?

Classroom numbers.

Thinning and prescribed burns cost $7 million, and each net benefit is already net of that cost. Have students compute one row each.

CBA says yes. Then ask: does the 1% branch bother you? It moves the expected value by only $1 million, yet it is a $100 million loss. That sets up the next slide and risk aversion.

🔥 When a Prescribed Burn Escapes

99.84% of Forest Service prescribed fires go as planned

341,000 acres burned when two escaped in New Mexico, 2022

$3.95 billion set aside by Congress to compensate victims

• Small probability, huge loss.
• Burns also add smoke now to reduce smoke later.
• Who bears the risk of the 1% branch?

The Hermit’s Peak and Calf Canyon fires merged in 2022. The Forest Service ignites thousands of prescribed fires a year, and escapes are rare but can be very destructive. This is the real-world version of the 1% branch. Answer to the question: neighbors bore it first, then federal taxpayers through the compensation fund Congress created.

😬 Risk Aversion

• A: $200 for sure
• B: coin flip for $400 or $0
• Both have expected value $200.
• Risk-averse people choose A.

What would you pay to avoid a small chance of losing your home?

Each extra dollar adds less well-being, so losing $200 hurts more than gaining $200 helps. A risk-neutral person is indifferent; a risk-loving person prefers B. The vertical gap is the well-being gained by removing the risk. This is why a household may pay a premium above its expected loss, which returns in the insurance choice at the end of class.

🎲 Risk vs. Uncertainty

🎲 Risk

We can list outcomes and trust the probabilities.

A hillside with a long fire record.

❓ Uncertainty

We cannot pin the probabilities down.

Extreme fire behavior in a hotter climate with more homes.

• Often we cannot even estimate the odds.
• Insurers’ own wildfire models disagree widely.

Under uncertainty, what replaces \(E[NB]\)?

Past fire records may be a poor guide when climate, fuels, and development are all changing. Insurers classify the same parcels very differently. Until 2024, California required rates based mostly on past losses, so this distinction returns in Class 3.

Answer: there is no single replacement. Options include a margin of safety, stress tests against a worst plausible case, and choosing actions that do reasonably well across many scenarios. The 2023 Canadian season, more than double the old record, is the kind of event that past records did not predict.

🏠 Do Building Codes Work?

• 39 of 100 exposed homes were destroyed on average.
• A California home built since 2008 is 13 points less likely to be lost than a 1990 home in the same fire.

The study compares homes with identical fire exposure but different building years and code rules, using U.S. homes exposed to wildfires since 2000. Thirteen points is 34% of the 39% sample average; the 39% is the sample mean, not the rate for a 1990 home. The grid is a picture of that arithmetic.

Codes for new homes pass a cost-benefit test at much lower fire probabilities than full retrofits of old homes, so where new homes get built matters.

🏘️ My Roof Protects Your House

• A pre-code neighbor within 10 meters raises your loss risk 1.7 points.
• Hardening has a positive externality.
• Private \(h_P\) falls short of \(h^*\).

Same logic as Canadian smoke. What is different here?

The 1.7 percentage points is a 4.4% increase. Fire spreads from structure to structure, so one owner’s choice affects the people next door. This is the externality logic from earlier in the course, with the sign reversed from pollution.

Answer: the neighbors live on one street under one local government. Bargaining, homeowners associations, and local codes can internalize the spillover; across a national border, none of those tools exist. Ask why neighbors still fail to bargain: many parties, free riding, and hard-to-verify work.

📝 Practice: Private and Social Returns

A $10,000 retrofit avoids $2,000 a year of expected damage to your home and $500 a year to your neighbors, for five years, at 4%.

Perspective	Present value	Net of $10,000
Owner only	$8,904	−$1,096
Owner and neighbors	$11,130	+$1,130

• What is the smallest grant that makes the owner say yes?
• Why might a mandate be cheaper than a grant?

Hide the table at first and let pairs compute. The five-year annuity factor at 4% is 4.4518: $2,000 × 4.4518 = $8,904 and $2,500 × 4.4518 = $11,130. The smallest grant is $1,096. A grant costs public funds and pays some owners who would have hardened anyway. A mandate avoids that, but it forces hardening on owners for whom it does not pay. Classroom numbers; link to present value from earlier in the course.

🏦 Sharing the Risk

The insurance choice ends Class 2. Class 3 (50 minutes): pooling and correlation 4, premiums and price ceilings 9, FAIR Plan 5, adverse selection 4, moral hazard and firefighting 8, underinsurance and biases 8, disclosure and recovery 6, town council 6.

🏡 Wildfire Insurance Choices

Your home is most of your wealth: a “one place-based egg.”

A 2% annual chance of fire. $400,000 to rebuild. Expected annual loss: $8,000.

Option	Annual premium	You pay if it burns
A: No insurance	$0	$400,000
B: Partial, subsidized	$5,000	$100,000
C: Full, risk-based	$9,500	$2,500

Which would you choose?

Classroom numbers.

Take a hand vote before showing the next slide. Numbers are set so each premium can be compared with its expected claim. A mortgage lender usually requires coverage, so option A is often not open to borrowers.

⚖️ What Each Option Does

• Expected yearly cost: A $8,000, B $7,000, C $9,550.
• C costs most on average and removes the ruinous year.
• Why is B cheaper than the risk it covers?

Exit ticket: Why might a risk-averse household rationally choose the option with the highest expected cost?

B’s expected claim is 2% × $300,000 = $6,000, above its $5,000 premium, so someone else pays the gap. C’s expected claim is 2% × $397,500 = $7,950; the $1,550 above that pays for administration, capital, and reinsurance. Exit-ticket answer: C removes the ruinous $400,000 year, and with diminishing marginal utility the certainty is worth more than the $1,550 loading.

🔗 When Claims Arrive Together

• 1,000 homes, $1 million each, 2% risk.
• Pooling works when losses are independent.
• Fire jumps roof to roof.

Classroom numbers.

Open Class 3 here. The top row uses a binomial distribution with 1,000 homes and a 2% chance; the 99% range is 10 to 32 claims. The bottom row is the limiting case. Same average, very different capital needs. Correlated losses need capital, geographic spread, or reinsurance, which is insurance for insurers. Actuaries must plan for the billion-dollar year so that one event does not bankrupt the firm.

📈 Premiums Rose Almost Everywhere

• Average California premium: +84%, or +45% after inflation.
• Still about 85% of the U.S. level.
• Should low-risk ZIP codes rise this much?

Premiums come from mortgage escrow records through March 2026, for policies from regular insurers. The 85% compares California with the rest of the country over the last ten years; lower California premiums may reflect regulation. Nationally, the pass-through of disaster risk into premiums has doubled since 2014, partly from repricing in global reinsurance markets, and higher premiums have slowed home price growth in the most exposed ZIP codes.

On the question: when premiums rise almost equally in every risk class, prices are not sorting risk. That is a clue for the next two slides.

🚧 Price Ceilings and Insurer Exit

• Proposition 103 (1988): rate hikes need prior approval.
• Until 2024, rates rested on past losses while risk rose.
• Below cost, insurers exit.

The argument: price suppression, not only a market failure, drives the retreat. Regulation also has a purpose: voters passed Proposition 103 after large rate increases, and it can protect consumers from excessive rates. In December 2024, California began allowing catastrophe models and reinsurance costs in rate-making, and required insurers using them to write more policies in distressed areas. Fixing availability this way creates new affordability pressure in high-risk areas. Ask for the Econ 101 prediction, then what the graph leaves out.

🏦 The FAIR Plan: Insurer of Last Resort

• From about 2% to 5% of single-family homes, 2020–2026.
• 2025: a $1 billion assessment on member insurers.
• Half can pass to policyholders. Who pays?

Data: California FAIR Plan.

Created in 1968, the plan is an association of all licensed property insurers that sells basic coverage to owners who cannot find it elsewhere. Exposure is the sum of policy limits; the 2021–2023 points are backed out of reported percentage changes. It was never designed to be the state’s main wildfire insurer.

Incidence: the February 2025 order bills member insurers, who may recover half through a temporary fee on policyholders. Statutory incidence falls on insurers; economic incidence can reach all policyholders, including low-risk ones, and insurers’ owners. Plan rates must be actuarially sound by law, so whether a given policy is subsidized is an empirical question.

🍋 Adverse Selection

• One price for all: lower-risk owners overpay.
• If they leave, the pool gets riskier and the price rises.
• Better risk models reverse the problem: insurers pick the safe homes.

Classroom numbers.

Adverse selection: at a given price, the buyers who stay are riskier than average. With equal numbers of each type, the pooled price of $3,000 breaks even only if both types stay. If lower-risk owners drop coverage or move to a cheaper insurer, the price must rise toward $5,000. Homeowners with mortgages cannot drop coverage, which slows the unraveling.

The flip side: insurers with parcel-level models offer low prices to safe homes, so firms with coarse models are left with the riskier customers in each price group. Winning a customer becomes bad news, and coarse firms raise prices in risky areas or stay out. Either way, the riskiest homes end up concentrated in the FAIR Plan.

🍪 Moral Hazard

Moral hazard: coverage changes behavior because you no longer bear the full cost.

• Insured owners get less from hardening.
• No “one-sided coins.”

“One-sided coins”: owners who keep the gains should not be able to pass the losses to others. Moral hazard requires that the insurer cannot price the action. If it can observe hardening and reward it with a discount, the gap closes. Insured owners still gain from avoiding disruption and uncovered losses, which is why the insured curve is lower but not zero.

🚒 Free Firefighting Is Free Insurance

• Expected protection can top 20% of home value.
• It rises with hazard and falls with density.
• Owners do not pay for it.

Data: NIFC, deflated with the CPI. Lines are period averages.

Real federal suppression spending rose by a factor of 3.4 between the two periods. It excludes state and local spending, and not all of it protects homes. Measured by how much nearby homes raise firefighting costs, the expected value of this protection is largest for a small number of high-risk, low-density homes. It works like subsidized insurance that never sends a bill. Who should pay: the landowner, the county, or all taxpayers?

🧠 Behavior and Recovery

About 20 minutes including the town council. If time is short, run the six reasons slide quickly and keep the town council.

🧩 The Opposite Problem: Too Little Coverage

• Buyers compare premiums and neglect coverage.
• 10 points more underinsurance: 4 points less likely to file to rebuild within a year.

Figure uses classroom numbers.

Evidence from Colorado contracts and households that lost homes in the 2021 Marshall Fire. Underinsurance means coverage falls short of the estimated rebuilding cost. Buyers treat insurers that write less coverage as if they had lower prices. Removing this coverage neglect would raise consumer surplus by about $290 a year, or 10% of premiums. Filing a permit is a step toward rebuilding, not a finished home.

🧠 Six Reasons We Underprepare

⏱️ Myopia

“The premium is due now. A fire is years away.”

🌤️ Optimism

“It won’t happen to my house.”

🛋️ Inertia

“I’ll deal with the brush next year.”

🧽 Amnesia

“We haven’t had a fire in years.”

🗺️ Simplification

“We’re not in the red zone.”

👥 Herding

“None of my neighbors did it.”

Which one is the hardest to fix with better information?

These six biases come from behavioral research on disaster preparation. Amnesia shows up in housing markets: prices near a fire dip and then recover within a few years as fear fades. Not every economist accepts this framing; one view holds that people respond sensibly to the rules, and that bailouts teach them not to prepare. Ask which view fits the next slide.

🏷️ Disclosure Moves Prices

• California sales, 2015–2022.
• Homes required to disclose sold about 4.3% less than nearby homes.
• The gap grew after damaging fires.

Figure is stylized.

The design compares homes near the edge of a disclosure zone and assumes other price determinants change smoothly at the boundary. This is hedonic pricing from Lecture 7. The discount bundles beliefs about fire, insurance costs, and the hassle of living with risk, so 4.3% is not a fire probability. Disclosure works on simplification: it turns a vague risk into a fact a buyer must read.

🏘️ Who Recovers?

26% of a year’s income lost over three years by occupants of destroyed homes

1 in 2 destroyed addresses still unoccupied four years later

• Burned areas are richer than average.
• Inside the perimeter, lower-income homes burn more often.
• New residents have higher incomes.

A rebuilt street, or recovered people: which do we measure?

Evidence from property damage in over 100 U.S. wildfires linked to tax and Census records. No income or employment effects appear for residents without direct property damage. A neighborhood can rebuild while its former residents do not return: places recover, people may not. Connect to smoke: those least able to defend against smoke are also least able to rebuild.

🏛️ Town Council Decision

A town has older homes near a forest, one road out, rising premiums, and smoke days most summers. It can fund one action before a fire and one after.

• Which losses does each action reduce or finance?
• Who benefits, and who pays?
• What would you measure to judge success?

Exit ticket: What can insurance do that hardening cannot, and what can hardening do that insurance cannot?

Give groups five minutes. Examples: targeted hardening grants, a second evacuation route, a fuel break with a smoke plan, clean-air rooms in schools and libraries, temporary housing aid, or relocation help. Require a measurable outcome. Exit-ticket answer: insurance moves money to the household after a loss, so it can pay for losses no one prevented. Hardening reduces the physical loss, including losses insurance never covers, such as lives, disruption, and neighbors’ homes.