Fully air-conditioned schools showed one fifth the learning loss in a July heat review

The sturdiest design in the pile belongs to the PSAT study, published in the American Economic Journal: Economic Policy. It pooled data from 10 million U.S. students between 2001 and 2014 and looked at the ones who sat the test more than once. For those students it compared a score in a year with hotter school days against the same student's score in a cooler year. That removes what is fixed about a student and a school: family income, teachers, curriculum, neighborhood. Anything else that changed for that student between sittings stays mixed in with the heat. It also points at cumulative exposure across a whole school year, not only the temperature on the day of the test.

The finding a school board will reach for is the one about cooling. "In our review, the most important finding was about air conditioning," Christine Ekenga, the Emory University epidemiologist who was senior author of the July review, said. In one of the included studies, "the schools that had full air conditioning experienced about one fifth the learning loss that was observed in schools without air conditioning," she said. Four fifths of the measured loss sits on the uncooled side of that comparison. But the comparison is across schools. Ekenga's work and earlier reporting also find the greatest learning loss among students from low-income families and from racially marginalized groups, and income is also what predicts which buildings have cooling. So other differences between those schools are counted in the one-fifth figure along with the cooling.

On why heat should touch a score at all, Ekenga pointed to several potential mechanisms. The most straightforward, she said, is the physiological strain associated with heat fatigue: it can make it harder to focus and maintain attention, and it can interfere with cognitive function. She also said thermal discomfort, the physical uncomfortableness of heat exposure, can make it difficult for students to pay attention.

Schools in the Northeast and the Pacific Northwest historically had no need for air conditioning, because classes began and ended outside the summer heat season. Where a city has acted, installation and function are separate questions: no rule requires New York's Department of Education to check that the units it put in are working, and the initiative left out spaces such as cafeterias.

Scientific American reports that putting air conditioning into older school buildings is not cheap, and it did not put a number on it. So the choice between a retrofit and a cheaper measure such as a later start date stays unpriced by this evidence. Ekenga put part of the fix outside the building. "Thermal comfort in school is just as important as what's happening outside of the schools and in the neighborhoods as well," she said.