Download the full report (pdf, 1.35 mb)
Executive Summary
Wildfires, which are increasing in frequency, duration, and intensity, are measurably affecting vulnerable populations, labor, housing, and education. This report describes how wildfire smoke disrupts various sectors of the economy across the United States. Wildfire smoke is a growing problem for groups that face greater economic barriers than the general population, such as low-income families, housing-vulnerable communities, and frontline workers.
Key Takeaways
- In the past decade, most Americans have experienced statistically significant increases in days of light, medium, and heavy wildfire smoke and decreases in smoke-free days.
- Increases in the number of days of smoke were greatest for the most dense, dangerous, and disruptive category of smoke.
- Avoiding wildfire smoke exposures is likely worth hundreds of billions of dollars per year to Americans.
- Increases in wildfire smoke are occurring in the nation’s most vulnerable communities, with disproportionate increases for minority populations and those with limited English proficiency.
- Frontline workers (here referring to those workers in outdoor occupations and often without indoor air filtration) are increasingly experiencing exposure to wildfire smoke. Smoke will continue to increase the risk of occupational hazards, decrease productivity, and cause worker disruptions in industries that depend on these workers. Adapting to these changing conditions will result in additional costs for businesses, consumers, and governments.
- Wildfires have increased heavy smoke exposures for young children and students in poverty. This could have impacts on early childhood and K‒12 education, such as disruptions in learning, poor academic outcomes, and increased food insecurity.
- Housing-vulnerable communities are experiencing an increase in heavy smoke days, especially in the high-cost regions of the West (The Federal Reserve’s Twelfth District). Wildfires are likely to pressure the housing sector by increasing housing costs and disproportionately impacting housing for vulnerable communities who live in housing types (older units, rental units, etc.) that are less likely to access protective adaptations.
- Dramatic increases in disruptive smoke overlap with eligibility for existing financing programs that could help build resilience to smoke-related damages. Programs that target low- and moderate-income communities and communities of color may have outsized importance in building broad economic resilience to climate risks.
Research Motivations and Methodology
Wildfire Smoke Health Effects
The health impacts of wildfire smoke exposure are not uniformly distributed across regions and populations. Certain populations, such as lower-income, children or older adults, medically compromised individuals or those who cannot avoid exposure, are especially vulnerable to smoke-induced health effects. Wildfire smoke exposure is associated with asthma exacerbations, chronic obstructive pulmonary disease, respiratory infections, myocardial infarction, ischemic heart disease, heart failure, dysrhythmia, pulmonary embolism, ischemic stroke, transient ischemic attack, out-of-hospital cardiac arrests, and all-cause mortality (Reid et al. 2016; Heaney et al. 2022; Wettstein et al. 2018). Such health outcomes as cardiovascular disease and cerebrovascular emergency department visits have been linked specifically to heavy-density smoke exposure, which has increased the most in the past decade (Wettstein et al. 2018). Previous research has shown that the negative health effects of prescribed fire smoke are more pronounced in children born to black and Hispanic mothers, as well as children of low-income mothers (Jones and Berrens 2021). Our findings suggest that individuals who are experiencing increased exposures might also live in communities with limited resources to reduce the impacts of the exposures.
Descriptive analyses were conducted on the presence of wildfire smoke plumes and their overlap with population centers to describe the magnitude of and trends in wildfire smoke affecting communities across the United States in 2011–2021. These data on census tract–level wildfire smoke exposures were combined with information on specific populations to characterize wildfire smoke exposures across different socioeconomic groups.
To describe recent trends in wildfire smoke, a comparison of estimates in the earliest five years (2011–2015) to those of the latest five years (2017–2021) of the 11-year study period was conducted. Using census tract aggregations of the daily smoke data, the mean annual days of smoke were calculated and then used to statistically test changes in frequency of wildfire smoke plumes across the study period. In each analysis, census tract estimates of person-days or number of smoke-days are used as the basis for central tendency estimates within the county or SVI (Social Vulnerability Index) tertile. All analyses were performed using R Statistical Software (R Core Team 2021).
Wildfire Smoke Exposures
To obtain community-level exposure to wildfire smoke, data from the National Oceanic and Atmospheric Administration (NOAA) Hazard Mapping System (HMS) smoke dataset were combined with population data from the 2010 U.S. Census. HMS data use satellite-detected fires with multiple daily satellite images and a combination of analyst examination and automated processing to record smoke plumes of categorical densities across North America. Satellite imagery that detects smoke plumes can reliably identify periods of wildland fire influence on ground-level measurements of air quality from validated monitors. Plume densities reported in HMS data correlate with PM2.5 concentrations, with concentrations <10 µg/m3 categorized as light, 10–21 µg/m3 as medium, and >21 µg/m3 as heavy.i
To estimate the sizes of populations potentially impacted by light, medium, and heavy wildfire smoke plumes between January 1, 2011, and December 31, 2021, smoke plume data and 2010 census block group centers of population were linked. Daily smoke density categories were assigned to populations in each block group if a smoke plume from any time in the day contained the block group population center. The spatial intersection of HMS plumes and population centers is detailed in Vargo 2020. Block group populations were held constant at 2010 levels to quantify the impact of changes in wildfire smoke regimes and disentangle them from population shifts over the course of the decade. Populations under each smoke category were considered for each day. The resulting quantity, person-days, is the product of the number of people in a census block group or tract and the number of days that block group experiences smoke. Person-days by smoke density and smoke-free person-days were then aggregated across geographies and time periods for our analyses.
After quantifying and describing general trends in wildfire smoke since 2011, the same data are combined with information on specific populations of interest to better understand who is most affected by wildfire exposures and how those communities might be prioritized for climate-resilient community development.
What is a person-day?
Throughout the report, person-days are used to capture, together, the number of people and the amount of time spent under smoke plumes. When a smoke plume is observed over a population center, each person who lives there is considered to have experienced one smoke day. Suppose 500 people live in a population center; each time a plume is over it, 500 person-days of smoke would be tallied. This measure can be adapted to consider communities of concern—for example, to count frontline worker—days, student-days, or household-days of smoke.
A person-day is a useful metric specifically because it incorporates people into descriptions of air quality. It helps to give an accounting of the potential impact of smoke by capturing the number of people and the amount of time people may have been exposed. Person-days assign exposures at fine scale but allow for versatile aggregation and comparison of exposures for different geographies and time periods.
Populations of Concern
There are several community dimensions of interest relevant to understanding wildfire smoke exposure and the resulting economic impacts. The characteristics of people or a community (e.g., age, race, health status, income, occupation), social inequalities (e.g., social capital, political power, lack of access to information), place-based inequalities (e.g., rural versus urban, elevation), and adaptation inequalities all impact a population’s susceptibility to disaster events and their resulting exposures (Cutter, Boruff. and Shirley 2003). Although wildfire smoke events affect entire populations together, their impacts are shaped by the population’s susceptibility and its adaptive capacity. This report’s findings suggest that increases in smoke are occurring in communities with high vulnerability in the labor, housing, and education sectors. Communities with fewer economic resources may face more barriers in avoiding exposures during a wildfire smoke event (Murphy et al. 2015). However, this overlap of vulnerability and growing exposure suggests that interventions that target at-risk communities may more efficiently reduce smoke exposure, potential health impacts, and social and economic losses associated with wildfires. This report is not exhaustive in its description of populations of concern. Considering other marginalized populations, such as indigenous communities, is important for improving understanding of the impacts of wildfire smoke.
The Centers for Disease Control and Prevention (CDC)’s SVI data were used to investigate populations of concern for wildfire smoke and evaluate characteristics that might affect the health risks of wildfire smoke exposures. All analyses were performed using the 2018 versionii of the SVI data at the census tract scale. Daily person-days of wildfire smoke at the block group level were aggregated to annual census tract aggregates and linked with 2018 SVI percentile rankings of four themes: (1) socioeconomic status, (2) race/ethnicity/language, (3) household composition and disability, and (4) housing/transportation. Estimates of person-days and number of smoke-days for each smoke density were calculated using national tertiles of the overall SVI theme and the four component themes. The tertile with the lowest SVI scores is referred to as having the greatest health/social “advantage,” and the tertile with the highest SVI scores is referred to as having the greatest health/social “disadvantage.” The assignment of tertiles using the census tract file (rather than other aggregations of SVI data) ensures that each tertile has roughly the same number of people. Additionally, specific components of the SVI (e.g., the number of persons without a high school diploma) were considered to examine changes in wildfire smoke among specific populations over the study period.
Frontline Workers
Wildfires have uneven impacts across the labor force and especially affect those who work outdoors or in indoor situations lacking adequate air conditioning or ventilation. Wildfire smoke impacts among these workers, referred to here as frontline workers, are expected to be greater than for other workers. Frontline workers are often paid lower wages, especially workers involved in food production and preparation or the movement and distribution of goods. These workers are also disproportionately racial and ethnic minorities. Migrant workers are also overrepresented in many of these frontline occupations, especially farmworkers and construction workers (Thomason and Bernhardt 2020). As a result of structural inequities, frontline workers have underlying health risks, low socioeconomic status, and reduced health-care access, which increases their overall vulnerability to wildfire smoke (Schenker et al. 2015). Frontline workers face increased occupational hazards, such as smoke-related health effects and exacerbated health vulnerabilities (Zhou et al. 2021), decreased productivity, and a greater likelihood of work disruptions and instability. As wildfire smoke increases hazards for these workers and disrupts productivity, the national economy suffers.
The contribution of industries to state labor forces and GDPs (gross domestic product), the percentage of workers considered frontline, and how frontline workers’ exposure to smoke changed from 2011–2015 to 2017–2021 were used to quantify smoke exposures in the labor force. The American Community Survey (ACS) five-year data from 2019 were used to enumerate frontline workers or those more likely to work outdoors and less likely to be able to mitigate their smoke exposures. Using estimates for variables included within the group C24050: Industry by Occupation for the Civilian Employed Population 16 Years and Over, the contribution of frontline industries to local labor forces and exposures among frontline workers were assessed. Among the 13 industries captured within the ACS group, frontline workers included in two occupations (“Natural resources, construction, and maintenance” and “Production, transportation, and material moving”) were counted as frontline workers and used with smoke days to arrive at frontline worker-days of smoke exposure. Four industries in the ACS variable with a majority of workers in frontline occupations were given special consideration: “Agriculture, forestry, fishing and hunting, and mining”, “Construction”, “Manufacturing”, and “Transportation and warehousing, and utilities”. State-specific GDP information was collected from the Bureau of Economic Analysis Table (SAGDP2N Gross domestic product by state) for year 2020.iii
School-Aged Children
The negative impacts on air quality make children a population of concern for wildfire smoke exposures. The development of the brain and organs throughout childhood and adolescence makes pollution potentially more damaging to children’s health, with much more long-lasting permanent effects, compared to adults (WHO 2005). Air pollution can decrease cognition and lead to poorer educational outcomes in the long term (Shier et al. 2019; Miller and Hui 2022). The economic impacts of childhood air pollution exposures can also impact near- and long-term school facilities and district budgets (Li and Jimenez 2022). A study of California schools from 2002–2003 through 2018–2019 found that wildfires related to nearly two-thirds of the school closure days and more than 70% of missed student-days over the 17 years (Miller and Hui 2022). Moreover, the study found significant negative impacts on academic performance among younger students. Another recent study found that the presence of wildfire smoke decreased students’ test scores, particularly for younger grades and disadvantaged districts. The impacts of one year, 2016, were projected to result in lost future earnings of more than $1.5 billion (Wen and Burke 2022).
Outside the classroom, school closures disrupt resources and services, such as meals and child care, and, as a result, students face increased risk of food insecurity and poor academic outcomes. Child-care providers and school districts with smaller budgets that serve disadvantaged communities could be particularly vulnerable to the impacts of wildfires because they often have greater needs and fewer resources available to address such issues. Parents are more likely to miss work to meet unexpected child-care needs, and households in low- and middle-income communities and communities of color are less likely to have reliable and affordable child-care options available to them (Harknett, Schneider, and Luhr 2022; Shrimali 2020). Additionally, child care disproportionately falls on low- and middle-income women and women of color, widening existing inequities (Shrimali 2020).
To estimate the impact of wildfire smoke on economically disadvantaged students, we used the ACS 2019 five-year estimates for the number of K–4 students enrolled in school and below the poverty line (variable group B14006) to calculate student-days of heavy smoke.
Housing-Vulnerable People and People Experiencing Homelessness
Affordable and safe housing is an important factor in dealing with many climate risks, as well as a commodity that is also threatened by climate risks. Homes are places of refuge from outdoor elements, such as wildfire smoke, and at the same time, fires pressure housing markets through loss of housing stock, limiting where new housing should be built, requiring retrofits, and increasing risks to existing properties. Lower-income residents will face disproportionate impacts due to the legacy of such practices as redlining, which segregated communities of color to neighborhoods that experience hotter temperatures and greater flood risks (Hoffman, Shandas, and Pendleton 2020; Katz 2021). Guidance for coping with wildfire smoke advises individuals to keep indoor air as clean as possible, often by closing windows and doors and running an air conditioner with a clean filter (CDC 2022). Lower-income households are more likely than higher-income households to live in housing that needs repair (Divringi et al. 2019), less likely to buy air filters, less likely to live in homes with air conditioning, and more likely to avoid running air conditioning due to the cost of energy (Hansen et al. 2011; English et al. 2007). Similarly, renters are more likely to be low‑income and thus more likely to rely on landlords to modify their homes to mitigate the effects of climate change.
Housing tenure is particularly important as a proxy for improvements and retrofits being installed, such as HVAC or window/door upgrades to control indoor climate. Few renters are in the position to invest in such improvements, and landlords are reluctant to pursue such investments (Melvin 2018). To consider the smoke exposures among housing-vulnerable households, estimates of renter-occupied households—along with cost-burdened (spending more than 30% of the household’s income on housing), owner-occupied households built prior to 1980—were counted for all U.S. census tracts.
The Comprehensive Housing Affordability Strategy (CHAS) data from the U.S. Department of Housing and Urban Development (HUD) were used to describe the overlap of smoke exposures with housing conditions that make it more difficult for residents to protect themselves (CHAS Database 2019). CHAS data are generated by HUD from custom tabulations of ACS data and provide estimates at the census tract scale within the range of the smoke record, specifically 2014‒2018. CHAS Table 12 was used to obtain detailed estimates of housing tenure (renter vs. owner-occupied), cost burden, year the structure was built, and household income. The estimates were combined with tract information about wildfire smoke exposure to describe household-days of smoke and changes across the study period.
People experiencing homelessness face a lack of regular shelter, as well as access to information and resources to prepare for and respond to wildfires, which amplify their wildfire smoke and health risk (Every et al. 2014; Gin et al. 2021; Gin et al. 2022). Additionally, many people experiencing homelessness are also working in low-wage, frontline jobs and thus represent a portion of the labor force especially vulnerable to disruptions from smoke exposures. A 2020 survey of people experiencing homelessness in Portland, Oregon, found that 75% did not receive any information during wildfires and 69% received no type of help during wildfire and smoke events (Hines, Petteni, and Knowlton 2021). Information on unhoused populations was obtained from HUD’s inventory of Point-in-Time (PiT) Counts assembled as part of the Annual Homeless Assessment Report to Congress (HUD 2021). Information about boundaries of Continuum of Care (CoC) were overlaid with census tracts to calculate the average number of days of smoke experienced by a CoC each year and to arrive at homeless-days of exposure. Although limited in their coverage, the PiT numbers give an idea of where the confluence of people experiencing homelessness and dangerous smoke resides.
Community Investment Opportunities
Several programs exist to direct investment to the communities where there are concentrations of disadvantaged populations discussed in this report. Given the intersectional nature of many of the factorsiv used to describe those disadvantages, programs targeting low- and moderate-income communities and communities of color can be important for fostering resilience to many climate risks and other social determinants of economic, physical, and mental health. The Community Reinvestment Act (CRA) is one example of legislation intended to ensure regulated banks help meet the credit needs of the local communities in which they operate. Specifically, banks are assessed on their record of meeting the credit needs of the entire community they serve, including low- and moderate-income neighborhoods. The U.S. Department of Housing and Urban Development (HUD) also identifies tracts for its Low-Income Housing Tax Credit (LIHTC) and those that are in Difficult Development Areas (DDAs) (CHAS Database 2019)—areas with high land, construction, and utility costs relative to the area median income and based on Fair Market Rents (FMRs) and income limits.
The Federal Financial Institutions Examination Council (FFIEC) identifies tracts for the Community Reinvestment Act (CRA). To be considered CRA-eligible, metropolitan tracts must be identified as low (tract median family income less than 50% of area median family income) or moderate (tract median family income greater than or equal to 50% and less than 80% of area median family income) income, or be identified as nonmetropolitan, middle (tract median family income greater than or equal to 80% and less than 120% of area family median income) income tracts designated by the FFIEC as distressed or underserved. Using CRA-eligible tract designations from 2020, changes in smoke exposures in qualified tracts were assessed.
Select Figures
The following figures are interactive versions of figures found in the full report.
Please review the related sections of the PDF (linked in each figure caption) for more discussion of the data.
National Trends in Wildfire Smoke, 2011–2021
Figure 1. Annual person-days of smoke exposure, 2011-2021 (billions)
Wildfire smoke exposures of all smoke densities have increased in recent years, with the largest increases in the most dangerous and disruptive category of smoke.
Wildfire Smoke Impacts on the Labor Force
Each industry’s share of the state labor force, the percentage of workers considered frontline, and how frontline workers’ exposure to smoke changed from 2011–2015 to 2017–2021.
Table 1. The total labor force, frontline workers, industry-specific workers, and frontline workers for each state are shown below.
| State | All Workers | Frontline Workers | % State GDP in Frontline-majority Industries* | % Increase in Frontline Worker-days Heavy Smoke |
|---|---|---|---|---|
| Alabama | 2,097,384 | 567,873 | 22.0 | 243 |
| Alaska | 347,774 | 84,622 | 29.0 | 277 |
| Arizona | 3,130,658 | 627,959 | 15.5 | 1121 |
| Arkansas | 1,303,490 | 363,291 | 22.8 | 102 |
| California | 18,591,241 | 3,912,179 | 17.0 | 2132 |
| Colorado | 2,904,589 | 554,940 | 13.4 | 534 |
| Connecticut | 1,815,636 | 316,882 | 15.3 | 266 |
| Delaware | 455,620 | 92,340 | 10.6 | 130 |
| District of Columbia | 376,871 | 25,382 | 1.6 | 125 |
| Florida | 9,495,353 | 1,885,667 | 10.1 | -16 |
| Georgia | 4,834,622 | 1,152,807 | 15.0 | 199 |
| Hawaii | 680,258 | 128,731 | 7.7 | |
| Idaho | 792,237 | 200,808 | 19.6 | 301 |
| Illinois | 6,250,862 | 1,389,558 | 18.2 | 232 |
| Indiana | 3,202,509 | 912,684 | 31.4 | 571 |
| Iowa | 1,613,902 | 432,703 | 26.5 | 230 |
| Kansas | 1,440,453 | 347,352 | 23.3 | 269 |
| Kentucky | 1,978,477 | 546,020 | 25.7 | 806 |
| Louisiana | 2,033,758 | 502,014 | 23.7 | 80 |
| Maine | 670,417 | 150,812 | 14.4 | 1021 |
| Maryland | 3,073,886 | 513,234 | 9.9 | 164 |
| Massachusetts | 3,612,375 | 579,690 | 12.2 | 276 |
| Michigan | 4,654,930 | 1,155,157 | 22.9 | 197 |
| Minnesota | 2,958,615 | 643,996 | 18.5 | 156 |
| Mississippi | 1,235,224 | 353,552 | 23.4 | 189 |
| Missouri | 2,916,000 | 687,662 | 17.5 | 271 |
| Montana | 512,329 | 117,394 | 17.1 | 259 |
| Nebraska | 999,212 | 243,424 | 24.7 | 254 |
| Nevada | 1,406,568 | 296,853 | 12.4 | 745 |
| New Hampshire | 729,701 | 152,236 | 13.9 | 393 |
| New Jersey | 4,422,491 | 823,212 | 13.7 | 227 |
| New Mexico | 888,646 | 184,770 | 15.4 | 143 |
| New York | 9,498,320 | 1,628,158 | 7.3 | 281 |
| North Carolina | 4,764,135 | 1,137,887 | 20.4 | 162 |
| North Dakota | 402,322 | 102,235 | 30.1 | 265 |
| Ohio | 5,595,444 | 1,367,433 | 21.0 | 581 |
| Oklahoma | 1,772,123 | 453,140 | 24.9 | 249 |
| Oregon | 1,979,043 | 427,407 | 19.5 | 332 |
| Pennsylvania | 6,199,456 | 1,400,992 | 18.4 | 369 |
| Rhode Island | 533,878 | 103,147 | 11.1 | 224 |
| South Carolina | 2,275,531 | 571,737 | 19.6 | 107 |
| South Dakota | 443,891 | 109,379 | 18.7 | 203 |
| Tennessee | 3,109,872 | 795,940 | 20.4 | 254 |
| Texas | 13,253,631 | 3,159,925 | 21.3 | 44 |
| Utah | 1,497,354 | 329,886 | 17.0 | 793 |
| Vermont | 329,028 | 68,847 | 14.6 | 400 |
| Virginia | 4,156,018 | 783,271 | 12.3 | 242 |
| Washington | 3,594,279 | 778,734 | 14.0 | 334 |
| West Virginia | 740,910 | 184,617 | 25.0 | 307 |
| Wisconsin | 2,982,359 | 782,642 | 23.5 | 199 |
| Wyoming | 288,503 | 82,594 | 32.2 | 489 |
| United States | 154,842,185 | 34,213,775 | 16.8 | 336 |
Source: American Community Survey (ACS) 2019, Industry by Occupation for the Civilian Employed Population 16 Years and Over; Bureau of Economic Analysis (SAGDP2N); Hazard Mapping System Smoke Product from National Oceanic and Atmospheric Administration (NOAA). For more details and analysis see pages 15-19 in the full report (pdf, 1.33 mb).
Wildfire Smoke Impacts on Children and Schools
Nationally, there were 569 million heavy smoke student-days (grades K–4), with 100 million among students below the poverty line (Figure 5). Between 2011–2015 and 2017–2021, heavy smoke student-days increased 300%.
Figure 5. Student-days of heavy smoke for all K-4 students (light) from 2011 to 2021; student-days for those in poverty (dark)v
Heavy smoke exposures among vulnerable students have increased dramatically in recent years. Younger students (grades K‒4), particularly those in poverty, are most affected academically by school closures.
Download figures (zip file, 14 mb)
Download data (zip file, 29 mb)
View data on wildfire smoke for your area.
The views expressed in this report are those of the authors and do not necessarily reflect the views of the Federal Reserve Bank of San Francisco or the Federal Reserve System.
Acknowledgments
Thank you to all the colleagues at the Federal Reserve Bank of San Francisco from community development, the web team, and communications who helped with this report. Additional thanks to Dr. Katie Conlon and Dr. Maria Mirabelli for their partnership in this work, and to the Climate Change and Health Equity Section at the California Department of Public Health.
End Notes
i. Particulate matter (PM) is a measure of air pollution that refers to inhalable particles made up of various chemicals. PM2.5 refers to particles that are generally 2.5 micrometers and smaller (Source: EPA Particulate Matter [PM] Basics).
ii. The 2018 version of the CDC’s Social Vulnerability Index is the most recent release at the time of the analysis. It is the case that neighborhood/population characteristics shift over time, such that a neighborhood’s SVI score in 2010 may be different than in 2018. By using the 2018 designations, the analysis highlights where smoke exposure changes over the last decade in neighborhoods that recently rank among the nation’s most vulnerable.
iii. Industry codes “11, 21”, “31-33”, and “22, 48-49” from the SAGDP2N data are used to capture industries with majority frontline workers.
iv. For examples of some of these “social determinants of health,” see the individual indicators of the CDC’s Social Vulnerability Index.
v. Reflects only changes in smoke over the study period and not changes to the number of students or students in poverty.
Article Citation
Lappe, Brooke, and Jason Vargo. 2022. “Disruptions from Wildfire Smoke: Vulnerabilities in Local Economies and Disadvantaged Communities in the U.S.” Federal Reserve Bank of San Francisco Community Development Research Brief 2022-06. doi: 10.24148/cdrb2022-06.