The Wildfire Economy Has More Than One Footprint
This series examines household economic trajectories and financial well-being after U.S. wildfires. Its central question is larger than how many structures burned: what happens to income, work, health, debt, and community membership when a fire destroys a home, when smoke reaches households that never see flames, and when some families have more financial protection than others? The evidence gives readers a way to distinguish those pathways and judge what each estimate actually shows.
That distinction matters because a wildfire has at least two economic footprints. The first is visible and concentrated: a fire burns property, displaces residents, interrupts local activity, and can damage the assets around which a household has organized its life. The second is broader and less visible: smoke and fine particulate pollution travel beyond the perimeter, bringing health risks, preventive expenses, work disruption, and pressure on household credit. These footprints overlap physically, but they should not be treated as the same exposure.
The difference is not just a matter of mapmaking. It changes who is counted as affected, which mechanisms are plausible, and how recovery is measured. A direct property loss can be followed through an address and a household's income record. A smoke shock may appear first as a medical concern, a missed work opportunity, a credit-card purchase, or a late payment in a place where no building was damaged. A community may also look more prosperous after rebuilding if the households who return or replace those who left have higher incomes, even while many original residents remain absent.
The map of flames is not the map of loss
The Census Bureau study at the center of the direct-damage evidence links property-level damage from more than 100 U.S. wildfires with tax and Census records. Its summary contains a finding that complicates any simple picture of who is vulnerable. Residents of burned areas are wealthier than the U.S. population on average, yet within wildfire perimeters the likelihood that a home burns is negatively correlated with pre-fire income. In other words, the average profile of an exposed place does not tell us how losses are distributed among the households inside it.
That distinction between an area and the people within it is essential. A high-income region can contain households with very different housing quality, mobility, insurance arrangements, and ability to finance a temporary move. Conversely, a household with a modest income can be exposed in a place whose average statistics look comfortable. “Who lived near the fire?” and “whose home was destroyed?” are therefore different questions. The second question is closer to the economic loss that follows a property event, while the first can conceal the distribution of that loss.
The same study reports that occupants of destroyed homes have lower income for three years after a fire. The cumulative loss is estimated at 26 percent of pre-fire annual income. That number is not a rebuilding invoice, and it is not a claim that every affected household loses precisely the same amount. It is evidence that the economic consequence of destruction extends beyond the physical cost of replacing a structure. Displacement, interrupted work, local disruption, and the difficulty of restoring a household's prior position can persist after the immediate flames are gone. The public summary does not decompose those causes, so the responsible conclusion is about a multi-year change in economic trajectory, not a single proven mechanism.
The distinction between an income trajectory and a one-time cost also changes how recovery should be judged. A grant, insurance payment, or rebuilt structure may address an immediate balance-sheet need without replacing the income that was not earned during disruption. The reported cumulative figure is useful precisely because it keeps time in view: the question is not only what a household lost on the day of the fire, but how its economic position evolves over several years. At the same time, the available summary does not tell us how much of the loss came from displacement, lost employment, household composition, or local economic conditions. It would be an error to treat the headline result as a fully separated causal account.
The study also reports no detected income or employment effects for residents without direct property damage in its sample. That finding is important because it sets a boundary: not every person living near a fire should automatically be described as having suffered the same direct economic loss as a destroyed-home occupant. But it does not make “not burned” equivalent to “unaffected.” A second study, using a different population and a different exposure definition, finds earnings and employment declines among borrowers exposed to higher levels of wildfire-related pollution outside the burn zone. The two findings need to be held apart rather than forced into a single average.
Recovery widens the distinction between a household and a place. Four years after a fire, half of destroyed addresses in the Census study remain unoccupied. Reoccupied addresses contain residents with substantially higher incomes than before. That pattern can coexist with new construction or higher average income among people living there, yet still represent an incomplete recovery for the households who were displaced. A repaired address and a restored livelihood are related outcomes, not interchangeable ones. Occupancy asks whether someone is back at an address; household income asks how an affected family is doing; neither measure alone captures the whole social and economic change.
Two routes from one wildfire
The Philadelphia Fed study makes the wider footprint visible by separating direct burn effects from smoke and pollution effects. For the main smoke analysis, it removes the burn zone and nearby areas, then compares households farther from the fire boundary that experienced heavy pollution with similar households in lightly polluted areas before and after the event. Its primary case is the 2018 Camp Fire, considered alongside three other extreme California wildfires from 2016 through 2020. The design is meant to ask a focused question: what changes when households are exposed to wildfire-attributable pollution without being inside the area of direct structural damage?
The comparison is deliberately more precise than “near a fire” versus “far from a fire.” In the main analysis, the study examines zip codes five to 30 miles from the boundary, generally treating average daily PM2.5 above 40 micrograms per cubic meter as heavy pollution and below 10 as light pollution. The immediate burn area is left out so that destroyed homes, evacuation, and other direct spillovers do not become part of the smoke estimate. The researchers then compare the same kinds of outcomes before and after the fire, using several pollution measures and detailed fixed effects to reduce the chance that ordinary differences between places explain the result. This is a quasi-experimental comparison, not a random assignment of smoke, so its strength lies in the design and robustness checks, not in pretending that all uncertainty disappears.
The answer is not that smoke produces a smaller version of property destruction. It creates a different chain of pressure. Pollution can worsen respiratory conditions and increase the need for protective measures. It can interrupt work or reduce productivity. Those changes can raise spending while weakening the income available for repayment. The financial record may then show higher credit-card and personal-loan delinquency even though the household never had to rebuild a burned home.
The study reports credit-card and personal-loan delinquency increases of several percentage points among households in the heavier-pollution group near the Camp Fire, with effects that do not disappear immediately. It also finds declines in earnings and employment in its labor-market analysis. Those are measured earnings and employment outcomes, not a complete estimate of permanent career damage. They nevertheless show why a smoke event can be economically consequential without a property claim: the route through the household may be health and work rather than a destroyed asset.
Distance does not end the story. The study finds smaller but still elevated credit effects 30 to 100 miles from the Camp Fire and reports related spending and repayment patterns across wider areas of Northern and Central California. The authors use a back-of-the-envelope calculation to illustrate what population scale can mean. Applying Camp Fire estimates to 19 million people in the New York metropolitan area exposed to Canadian wildfire smoke produces an imputed $6 billion in additional credit-card spending and $10 billion in added credit-card debt. Those figures are not observed New York totals; they depend on transferring an estimate from one setting, adjusting for exposure duration, and assuming comparable effects. Their value is conceptual: a modest effect per household can become economically large when the treated population is dispersed and enormous.
The two routes can also overlap without becoming analytically identical. A household inside a fire perimeter may breathe smoke, lose work, and face property damage at the same time. The smoke analysis removes the burn zone precisely because it is trying to isolate the portion that can affect distant households. The direct-burn analysis, by contrast, compares people inside the footprint with nearby people outside it who may still share some smoke exposure. Each design is answering a different question. Reading their estimates as though they were competing measurements of one treatment would erase the very distinction that makes the evidence useful.
How to read the evidence
The two studies provide a strong starting model, but they do not constitute one harmonized experiment. The Census paper follows property damage and administrative outcomes across a broad set of U.S. wildfires, while the Philadelphia Fed paper uses detailed credit, pollution, health, and workforce data concentrated in selected extreme California fires before the COVID-19 pandemic. Their samples, time horizons, outcomes, and research designs differ. Agreement between them is therefore informative, but it is not the same as replication of one estimate.
The most useful reading habit is to ask four questions of every claim. First, what was the exposure: a destroyed home, a location inside the burn footprint, or a measured level of smoke? Second, what was the unit of outcome: annual income, a credit account, employment, an address, or the composition of reoccupied homes? Third, what mechanism is directly measured, and what is only a plausible explanation? Finally, how far can the result travel beyond the studied fires and populations?
The unit matters because a percentage-point increase in delinquency, a percentage decline in earnings, a dollar change in annualized credit-card balance, and a vacant address describe different parts of the same event. They cannot be added together as if they were one loss measure. Nor can a higher average income among reoccupants be used to infer that the original residents recovered. A careful account keeps the denominator and time horizon attached to each finding.
This discipline prevents two opposite mistakes. One is undercounting wildfire harm by looking only at burned structures. The other is overgeneralizing by treating every smoke-exposed or nearby household as if it experienced the same loss. The evidence instead points to a layered geography of risk. Direct destruction can produce persistent income loss and severe financial distress. Smoke can affect a much wider population through health, work, spending, repayment, and delinquency. Insurance, tenure, and available credit can alter how the same shock appears in the financial record. And recovery can change the population of a place without restoring the people who lived there before.
The evidence also requires a distinction between what a study observes and what its authors infer. Credit-card spending and repayment are observed account behaviors; the exact purchase, illness, or work interruption behind each dollar is not. Health indicators and employment measures support a plausible route from smoke to financial stress, but they do not by themselves allocate every dollar of added debt to one channel. Insurance may explain why some direct-burn homeowners pay down balances while renters become delinquent, yet the Philadelphia Fed study does not observe the relevant claims directly. These qualifications narrow the claims without making the underlying patterns unimportant.
The clearest place to begin is the most visible pathway: what the data reveal when a home is destroyed, when a household remains inside a burn zone, and when the financial records of homeowners and renters begin to diverge. That baseline makes the later smoke evidence more precise—not because distant exposure is secondary, but because it shows exactly what the wider footprint adds to the economic story.