Students work in the classroom of a primary school in Lyon, September 2, 2021
JEFF PACHOUD / AFP
About the usefulness of air purifiers… even without Covid
Classes where the air is filtered would thus perform 20% better on exams than classes that are not equipped. The equivalent of a massive reduction in the number of students per class. Unfortunately, the Covid will not have been enough to convince the government to act decisively in this matter…
Atlantico: In 2020 you published a study titled Air filters, pollution and student performance. How important is good air quality and the installation of air purifiers for school performance?
Michael Gilraine: The correlation I found was quite large: installing air filters in classrooms increased test scores by 0.20 standard deviations. To put that number into perspective, this increase in test scores is about the same as that seen in a (pretty well-known) RCT conducted in the late 1980s called Project STAR, which reduced classes by (approximately) 22 to 15 students. Another way to assess the magnitude of the effect is to express it in months of learning. The gain of 0.20 standard deviation would correspond to approximately two and a half months of additional learning.
Are these results consistent with other similar studies?
Yes. Over the past five years, much research has been published showing that air pollution is detrimental to student achievement. I would like to express my appreciation for the excellent work Sefi Roth and Claudia Persico have done in this area. The effect sizes I find are about similar (a little more towards the higher end) to those in these other articles. I think it is now very clear that poor air quality reduces student performance on exams.
To what extent is this also a cost-effective way to improve the level of students?
It’s incredibly profitable. Air filters cost only about $700, and the cost per classroom to improve student performance by 0.2 standard deviations therefore compares very favorably with other interventions. If we factor in other costs such as depreciation, electricity, filter replacement, and installation of air filters in common areas, we arrive at a cost of approximately $1,000 per school year. This suggests a benefit-cost ratio of 0.2 standard deviation of test scores per $1,000 of expenditure, which would indicate that installing air filters is one of the most cost-effective educational interventions (in terms of dollars per test score) available to policy makers. Obviously, the effect of the air filter will likely vary widely from school to school (for example, the effects are likely to be lower in less polluted cities or in newer schools with better HVAC systems), but it appears to be very expensive nonetheless. effective compared to other educational interventions.
Some works, notably R. Jisung Park’s, have explored the usefulness of installing air conditioners by showing that high temperatures reduce performance. How important is a cool environment to study? Do you think it’s a continuum of things that makes for a good study environment?
I’m familiar with Park’s work on heat and learning, and it seems to be another channel that can affect student performance, especially since tests in the United States are often given in May when it can get quite hot. Overall, this highlights the importance of school environments, both in terms of air quality and temperature.
With the Covid crisis, the importance of filtering and renewing the air in closed spaces has become more important than ever. However, only a few installations have been made in Europe, and especially in France. Did we miss the opportunity to make a gesture that is both healthy and educational?
I have no expertise on the impact of air filters on virus filtration (and the impact of this filtration on the spread of viruses). I think air filters are a valid intervention in the absence of COVID, so to the extent that COVID encourages the installation of air filters, I see that as a good thing regardless of their impact on viruses.