How Climate Change Disrupts Kids’ Education

In the spring of 2022, the Calf Canyon/Hermit’s Creek fire burned over 300,000 acres in New Mexico¹, including terrain near the Rio Gallinas School for Ecology and the Arts in Las Vegas, New Mexico. “If you were in the parking lot, you could see all this smoke coming up,” says the school’s director, Aja Currey.

Las Vegas, New Mexico is a rural community—vastly different from its Nevada namesake—and children aged 5 to 14 travel from around the region to attend the school. Some students and staff evacuated from areas that burned and wondered if they’d have a home to come back to. Classes were in session for about two of the three months the fire raged, and children showed up when they could. But Currey estimates that around a quarter of the student body was absent at any given time. More recently, fires around Los Angeles have caused over half a million students to miss school.²

With climate change exacerbating fires and extreme weather events, the effects on education could be long-lasting. Recent research suggests that when children miss school or attend classes during hazardous conditions, learning setbacks linger long after the immediate crisis comes to an end. In some cases, a natural disaster can even force students to end their education prematurely. “There’s going to be more of [these disruptions] in the future,” Currey says. Researchers are just starting to elucidate how deeply kids’ educations will suffer—and how to mitigate the effects.

A Widespread Problem

In 2024, one of seven students worldwide had their education disrupted by the types of extreme weather made worse by climate change, according to a report from UNICEF.³ In Vietnam, for example, eight schools were completely destroyed or heavily damaged after a typhoon. In Africa, UNICEF estimates that 20 million children are at risk of dropping out of school because of climate-related disruptions that happened last year. And in the southeastern United States, tens of thousands of children missed weeks of school in the wake of Hurricane Helene. Globally, roughly 242 million students attended school for fewer hours than normal due to such extreme weather events.

UNICEF’s report is one of the first to quantify these impacts, making it difficult to determine whether climate-related school closures are on the rise. “There’s a huge lack of data,” says climate impact specialist Krishna Krishnamurthy of the Green Climate Fund in Incheon City, South Korea, who has previously worked with UNICEF to study how climate change affects education.⁴ But some school administrators say that today’s weather poses problems they rarely had to contend with in the past.

Take, for example, superintendent Matt Foster’s school district in northeastern Vermont. Heavy rain has caused the region to flood twice in recent years. Foster is used to snow days, but these typically have a minimal impact on education. “You shovel it, you move on,” he says. Destructive storms are a different beast, requiring debris removal and reconstruction that can take weeks.

Careful Calculations

Economists are now beginning to quantify how these events disrupt children’s education and how long the impacts last. One method involves examining how large numbers of students perform on standardized tests and comparing those scores in years with or without extreme weather events. That enables researchers to link these events to educational outcomes, irrespective of other aspects of the students’ lives. Another approach compares the individual scores of children who repeatedly take the same standardized test over several years.

In some cases, average test scores may only drop by a tiny fraction after an event. But because they affect so many students, these seemingly small effects can have a large cumulative impact, says economist Isaac Opper from RAND in Santa Monica, California. Educational attainment is linked to salary, and natural disasters can, in principle, cause groups of kids to miss out on cumulative lifetime earnings worth billions of dollars, which can exacerbate social inequalities.

When a child’s education is interrupted multiple times by extreme weather, the compounding effect can even lead them to leave school early, adds education researcher Caitlin Prentice from the University of Oslo in Norway.

When Storms Close Schools

Opper and colleagues have found that natural disasters like hurricanes can reduce test scores by up to 2% of the range between average- and high-scoring students, an impact similar to a young teen losing about 10% of the reading skills they would typically gain over a school year.⁵ Very large disasters, such as Hurricane Katrina, caused test scores to drop almost four times as much as a smaller disaster, such as a tornado that only caused localized damage.

Even more worrisome is that students’ test scores don’t seem to bounce back in subsequent years. For some children, the hit may be enough to end their educational careers early—full-time college attendance rates drop by about 2.6% among kids caught up in large disasters, Opper and his colleagues found. Without a college degree, each of those children are likely to earn at least $630,000 less over the course of their lives, according to the Social Security Administration.

In fact, lost lifetime earnings are comparable to the cost of rebuilding damaged infrastructure after a storm, according to the findings from Opper and colleagues. For example, if a city pays $1 million to repair the damage caused by a flood, all the students who missed school because of that flood will likely earn a combined total of $1 million less over their lifetimes.

These effects are particularly acute for low-income communities, where schools have the fewest resources to help children catch up, adds Cassandra R. Davis, who studies inequities enhanced by hazards and disasters at the University of North Carolina at Chapel Hill. In Vietnam, for example, heavy rainfall frequently prevents many students from attending school—so much so that 28% of the ethnic disparity in upper-secondary school enrollment can be can be attributed to hazards posed by rain. That’s according to a recent study based on weather observations collected by NASA and data from a study on socioeconomic issues in rural parts of southeast Asia.⁶

An obvious solution is for children to make up missed class time during weekends or school holidays. But in some cases, children simply miss too much school for this to be a viable option—some schools in Pakistan closed for six months in 2022 because of flooding.⁷ In the United States, budgets and district policies often prevent schools from extending their teaching hours, and teachers are rarely offered guidance on how to adjust their curricula to compensate for lost time, Davis says.

“Missing a month of the second grade is not nothing,” says Susanna Joy Smith, a writer and parent in Asheville, North Carolina, whose second-grader missed school for a month in the wake of Hurricane Helene last fall. Students there will only make up one-and-a-half days. The long shutdown amounts to “asking our children to really shoulder the burden of this disaster and pay for it in their educations,” Smith says.

In any case, making up time won’t necessarily address the damage done. Storms can leave psychological scars, and droughts can ruin harvests and leave kids malnourished, making it hard for them to focus on schoolwork. Likewise, heatwaves can make it hard for children to sleep, and lack of sleep has been shown to lower test scores.⁸ “Simply adding more lesson time, even if it was practically feasible, would not fully address the issue,” Prentice says

Smoke and Sweat

Parents, schools, and communities also need to worry about direct health impacts, such as toxic pollutants from wildfires. By analyzing standardized test scores and satellite data, environmental economist Marshall Burke at Stanford University in California, and his colleague Jeff Wen found that if wildfire smoke added 10 micrograms of particulate matter to the air during school days, test scores went down at the end of the year. In contrast, smoky weekends had little influence on test scores, suggesting a genuine correlation between classroom air quality and educational outcomes.⁹

Although the impact on test scores was small—an average drop of 0.029% of the usual range between high- and average-scoring students—the researchers calculated that the cumulative drag on education could result in a total loss of earnings running into billions of dollars. Burke also notes that they used datasets that stopped at 2016, and wildfire smoke exposure has significantly worsened since then.

The educational impact of climate change is not limited to discrete, extreme events. As global average temperatures rise, classroom conditions sometimes soar above 80°F, creating uncomfortable conditions that make it harder to learn.^{10 11}

To study the effects of overheated classrooms, economist R. Jisung Park from the University of Pennsylvania in Philadelphia tracked students who took the Preliminary Scholastic Aptitude Test (PSAT) multiple times over several years. Park found that if a school year was, on average, 1°F hotter, it reduced the expected gain in score upon retaking the test by about 1%.¹² This effect was much smaller when the school used air conditioning in its classrooms.

These small average effects could mask bigger problems for some students, says indoor air quality researcher Pawel Wargocki from the Technical University of Denmark. Children with learning disabilities, such as attention-deficit/hyperactivity disorder, are likely to be harder hit than kids who perform well even when conditions aren’t perfect. “In the end, average effects will be small, but disparities huge,” he says.

Park and his colleagues also looked at student scores on a standardized test called the Programme for International Student Assessment (PISA) across dozens of countries, some much hotter than others.¹³ Each day above 80°F in the three years leading up to the test lowered scores by an average of 0.15% of the normal range between average- and high-scoring students, they found. That suggests that greater heat exposure may explain up to one-third of the difference between PISA performance in hot countries, such as Brazil, compared to cooler countries, like South Korea.

It’s not actually easy to determine the optimal classroom temperature, Wargocki notes. People who live in warm climates are physiologically adapted to warmer rooms than people who live in cool climates. What’s more, children seem to learn better at lower temperatures than adults, probably because children have faster metabolisms and are more active. But, in general, cooler classrooms are best for learning since it’s easy to adapt by putting on more clothes.

Taking Action

As global carbon emissions continue to rise, schools will need climate-proofing strategies. To combat extreme heat without the aid of air conditioning, cities in Southeast Asia often use long buildings with strategically positioned windows and doors that foster cross-breezes. To avoid long school gaps due to flooding, a nonprofit organization in Bangladesh even operates floating schools that serve flood-prone areas, where it’s often challenging for students to access land-based schools. When floods hit, students can keep studying on boats equipped with classrooms, and even playground equipment on their decks.¹⁴ “I think it’s worked fantastically,” Krishnamurthy says.

In some regions, providing financial assistance for families could reduce the impact of heat on educational performance. Economist Teevrat Garg at the University of California San Diego studied an agrarian part of India where extreme heat has reduced crop yields and curtailed household incomes, leaving some parents unable to buy their children textbooks or even nutritious food. But when parents enrolled in a program that guaranteed them paid work, hot days were only about 50% as detrimental to their children’s test scores, presumably because parents used some of those funds to buy the items their children needed to be successful students, Garg found.¹⁵

And remote learning programs could also play a role, at least for short periods. They do tend to work best when parents are able to ensure that their children participate, Krishnamurthy says. During the COVID-19 pandemic, for example, one-on-one phone tutorials in Uganda offered an effective way to keep children learning.¹⁶ But remote learning during the pandemic had low participation rates in many other parts of the world and rarely matched the effectiveness of in-person learning, according to a World Bank report. That’s partly because many students lacked reliable access to technology or consistent parental support—both vital factors for successful remote learning.¹⁷ (See also “Kids adopt different ways of coping during the pandemic“.

In Asheville, kids are back in school in the wake of Hurricane Helene. But Smith worries that the storm will have a lasting learning impact on her son, who has dyslexia. And she fears the impacts yet to come. “What happened during Hurricane Helene is the new norm,” Smith says. And when it comes to education, “We as communities are fundamentally unprepared at so many levels for climate change.”

1. Natural Resources Conservation Service, NRCS New Mexico’s Hermit’s Peak/Calf Canyon disaster assistance (2024). https://www.nrcs.usda.gov/conservation-basics/conservation-by-state/new-mexico/nrcs-new-mexicos-hermits-peakcalf-canyon. Accessed 19 December 2024. ↩︎
2. G. Madakumbura et al., Climate change a factor in unprecedented LA fires (2025). https://sustainablela.ucla.edu/2025lawildfires. Accessed 19 December 2024. ↩︎
3. UNICEF, Learning interrupted (2025). https://www.unicef.org/reports/learning-interrupted-global-snapshot-2024. Accessed 19 December 2024. ↩︎
4. K. Krishnamurthy et al., It is getting hot (2019). https://www.unicef.org/eap/reports/it-getting-hot. Accessed 19 December 2024. ↩︎
5. I. M. Opper, R. J. Park, L. Husted, The effect of natural disasters on human capital in the United States. Nat. Hum. Behav. 7, 1442–1453 (2023). ↩︎
6. T. Pham, The child education and health ethnic inequality consequences of climate shocks in Vietnam. Econ. Educ. Rev. 90, 102311 (2022). ↩︎
7. M. Venegas, L. Schwarz, S. Sabarwal, “The impact of climate change on education and what to do about it” (Tech. Rep., World Bank, Washington, DC, 2024). https://openknowledge.worldbank.org/entities/publication/761bcf7b-4bb6-4fea-b75b-146603d437bb ↩︎
8. M. Jagnani, Children’s sleep and human capital production. Rev. Econ. Stat. 106, 983–996 (2024). ↩︎
9. J. Wen, M. Burke, Lower test scores from wildfire smoke exposure. Nat. Sustainability 5, 947–955 (2022). ↩︎
10. C. M. Prentice, F. Vergunst, K. Minor, H. L. Berry, Education outcomes in the era of global climate change. Nat. Climate Change 14, 214–224 (2024). ↩︎
11. P. Wargocki, J. A. Porras-Salazar, S. Contreras-Espinoza, The relationship between classroom temperature and children’s performance in school. Building Environ. 157, 197–204 (2019). ↩︎
12. R. J. Park, J. Goodman, M. Hurwitz, J. Smith, Heat and learning. Am. Econ. J. Econ. Policy 12, 306–339 (2020). ↩︎
13. R. J. Park, A. P. Behrer, J. Goodman, Learning is inhibited by heat exposure, both internationally and within the United States. Nat. Hum. Behav. 5, 19–27 (2021). ↩︎
14. J. Beaubien, ‘Floating schools’ make sure kids get to class when the water rises (2018). https://www.npr.org/sections/goatsandsoda/2018/09/12/646378073/floating-schools-make-sure-kids-get-to-class-when-the-water-rises. Accessed 19 December 2024. ↩︎
15. T. Garg, M. Jagnani, V. Taraz, Temperature and human capital in India. J. Assoc. Environ. Resource Economists 7, 1113–1150 (2020). ↩︎
16. N. Angrist et al., “Building resilient education systems: Evidence from large-scale randomized trials in five countries” (NBER Working Paper 31208, National Bureau of Economic Research, 2024). http://www.nber.org/papers/w31208 ↩︎
17. A. Muñoz-Najar et al., “Remote learning during COVID-19: Lessons from today, principles for tomorrow” (Tech. Rep., World Bank, 2021; https://www.worldbank.org/en/topic/edutech/brief/how-countries-are-using-edtech-to-support-remote-learning-during-the-covid-19-pandemic). ↩︎