The school year has barely started in Denver, and French teacher Tiffany Choi is already worried that her students are suffering from distraction.
The issue isn’t texting, watching TikTok videos or passing notes. It’s Denver’s ongoing heat wave.
“Today was a little bit hot, so I noticed kids were very sleepy and they were having to get up to drink water quite often,” said Choi, who works at Denver’s East High School. “If you are dehydrated, and you have to keep going to the water fountain, that can take away from their classroom experience.”
While nodding off in class on a warm day may seem like a right of passage for the average teen, Choi’s observation carries a bigger consequence than parched lips.
Extreme heat lowers a child’s ability to learn, according to an examination of 10 million American students that shows hotter school days reduce standardized test scores. Based on the analysis, every 1-degree-Fahrenheit increase in average outdoor temperature over a school year reduces student learning by 1 percent.
“There’s been quite a few media reports and anecdotal reports about teachers noticing that students weren’t able to focus on hotter days,” said R. Jisung Park, an environmental and labor economist at UCLA’s department of Public Policy and the Luskin Center for Innovation who co-led the analysis in a forthcoming paper* in American Economic Journal: Economic Policy. “My co-authors and I were looking to ask the question of, ‘does a hotter climate during the school year actually affect the rate of learning?”
That’s exactly what they found. The drops in academic achievement couldn’t be explained by hotter weekends or hotter summers, but the trend was connected to higher temperatures on school days alone.
One obvious remedy is air conditioning, so the team conducted what would become the first nationwide assessment of A/C access at schools. They found air conditioning could almost completely reverse the academic damage caused by warming. But paying to install air conditioning would cost many of these already-strapped school districts millions of dollars, the study estimates.
The connection between lost learning and a greater number of hot days is one more example of how climate change is already affecting our lives — and it’s an alarm bell for what we stand to lose in the future. Humans still have time to allay the worst consequences of continued global warming. But unless significant changes occur in the next decade — which seem more and more unlikely — the world will be locked into an inescapable period of heat waves unlike our species has ever seen.
Aside from glimpsing what this limbo period will be mean for young students and their ability to learn, the study adds to hefty debate around how developed nations will be influenced by global warming. The prominent theory is low-income countries will bear the brunt of consequences, but this study and others point toward nuanced calamities for places like the U.S.
What the economists did
This economic investigation was driven by a partnership with the College Board, the nonprofit that administers tests for higher learning, such as Advanced Placement exams, the SAT and its precursor, the PSAT.
The advantage of studying the PSAT is that students tend to take the test more than once — typically in 9th, 10th or 11th grade. The researchers collected PSAT scores for 10 million students who took the exam at least twice between 1998 and 2012 — for a grand total of 21 million test scores.
They then compared those test scores to daily data from 3,000 weather stations for the same time period, as well as data on local pollution and economic conditions. Those components were essential because it meant the team could track how individual students and school districts evolved alongside global warming — while also accounting for other things that might affect academics, like income levels.
“It’s almost like nature is the one running the experiment,” Park said. “And then we’re lining the data up in such a way that we can make stronger causal claims about what hotter temperature is or is not doing to learning.”
What the economists found
Notice that Park said “causal claims” and not “correlation.”
By amassing so much data, the team could directly assess if weather causes changes in learning — through the use of statistical methods called econometrics.
“This is a well-researched and trustworthy [study] design for understanding the causal effects of changes in things like temperature on economic outcomes,” said Marshall Burke, a Stanford University economist who wasn’t involved with the study. “So you take a school in New York and you don’t compare it to a school in San Francisco. You compare it to itself over time as the temperature fluctuates.”
Hotter temperatures on school days were the only weather marker that stood out. Even a school’s number of snow days had little to no impact on academic achievement.
“Hot weekends or hot summer vacation days or hot spring break days don’t seem to affect learning nearly as much as hot weekdays during the school year,” Park said. Beginning classes later in the fall likely wouldn’t help, as the study found hot days at both the start and end of the school year affected learning.
The Dixie divide
The sweet spot for academic improvement on the PSAT seemed to be class days that hovered around 60 to 70 degrees Fahrenheit, according to the study.
By contrast, every time a student experienced at least six school days above 90 degrees Fahrenheit, it reduced their yearly learning — or how much an average student would be expected to improve academically in a single year — by 1 percent. And these deficits appear to compound over time — students who had experienced hot school days two, three and four years prior to taking a PSAT had substantially lower scores.
Those temperature trends carved out a North-South divide on the U.S. map.
Hotter Southern counties — like in Florida and Texas — showed lower test scores than counties in the North, even after controlling for other socioeconomic factors like family income, county economic status or local pollution.
This geographical divide in school test scores has long been known. Researchers have blamed it on resource disparities and public-versus-private educational approaches dating back to the Civil War. But this study shows heat plays a role, too, and further affects groups who are already contending with systemic disadvantages.
“The causal effect of any given 90-degree day was much larger for lower-income students and racial minorities,” Park said. “And we know that those types of students are more likely to live in the South or in hotter parts of the country.”
Why does it matter — and can people fix it?
All of these findings serve as a warning for the mounting climate crisis.
“If we have two degrees of warming without any changes in infrastructure, then the implication is that the average student in the United States would learn around 7 percent less than he or she otherwise would have,” Park said. “Los Angeles already sees five to 10 more 90-degree days per year than it used to 20 years ago.”
So far this school year, extreme heat led to class cancellations in 50 Baltimore schools. Teachers in Madison, Wisconsin, and Oakland, California, are shutting off classroom lights to cope with high temperatures. Kentucky schools suspended all outdoor activities, fearing threats to children’s health like heat stroke.
Frances Moore, an environmental economist and climate scientist, said this study contributes to an ongoing debate on global warming’s effects on people’s productivity, especially in developed countries. Research has already showed that hotter temperatures in an office can make brains lethargic over the course of a single day — but now we know these effects can start early in life and intensify with time.
“Up until now we haven’t had a ton of evidence on the mechanisms by which climate change might have these long run effects,” said Moore, who works at the University of California, Davis and wasn’t involved with the study. “If you affect kids in this early stage of their life when they’re building this knowledge base, that’s going to have very long run implications on human capital.”
Access to air conditioning was one variable that seemed to reverse the course on these heat-related learning losses. For the average student, school air conditioning offset 73 percent of the learning reductions caused by hot days, the study found. Include air conditioning at a student’s home too, and the achievement gap essentially disappeared.
Universal access to A/C would obviously consume energy — and by extension, possibly lead to more carbon emissions. But the study stated that the A/C effect could also be due to certain schools having more resources for cooling overall, such as an ability to plant more trees on campus.
“The solutions are really tricky,” Burke said. “We do want air conditioning, right? We just want its energy to be generated from renewable sources.”
The costs for adapting schools will likely be hefty. According to Park’s study, an urban area like Houston would need to spend $2,100 per year per student — or $2.1 million per year for each high school with 1,000 students — for enough air conditioning to waylay heat effects on learning.
While almost all guidance counselors in the South reported their schools had air conditioning, the majority of schools in the Northeast lack it. This suggests that the regional gap in academic achievement would have been wider if not for the South’s cooling systems.
But it also means that the North and other historically cool places may be less prepared to deal with future warming.
That may be the case in the alpine city of Denver, where 60 of the city’s 207 schools operate without full air conditioning. East High School building, where Choi works, has stood since 1924 and lacked air conditioning until it was installed last year.
Prior to its installation, Choi says there was a noticeable difference in the past in how her eighth period classes performed on sizzling August afternoons, versus her classes that met first thing in the morning.
“I found it especially difficult for the freshman students to focus. They’re tired from being in school all day, but then it’s just sort of [amplified] by it being so hot,” Choi said. She said heat affected her too, making it harder to focus on her lesson plan and leaving her drenched in sweat at the end of the day. She tried opening the windows, but bees would fly in. At one point, she crowdfunded $800 to buy an A/C unit for her class.
Even with air conditioning, she said temperatures in the classrooms still feel like they reach the 80s — speaking to the notorious problem of artificial temperature control for very old buildings. And Park’s study confirmed a lot of what she suspected.
“That’s really interesting that they were able to track this with data, but it doesn’t surprise me at all. I’ve always thought this was a huge equity issue,” Choi said. “It’s much easier to think in a room that is cold than a room that is hot.”
*Editor’s note: An early version of the study appeared in the National Bureau of Economic Research in May 2018.