Readers may remember that a year ago, I spoke with researcher and AEI colleague Nat Malkus about his widely used Return to Learn (R2L) Tracker, which tracked in-person learning across 8,500 school districts. The R2L data were used by the White House and U.S. Department of Education in weekly briefings for senior staff, as well as by the CDC, Federal Reserve, and state leaders and an array of major media. Using the same innovative web-scraping technology, Nat has just released a and tool that makes it possible to analyze masking requirements across more than 8,000 school districts during the whole of the 2021-22 school year. As mandatory masking recedes, this provides an important opportunity to look back and ask what we鈥檝e learned. I was curious to hear Nat鈥檚 thoughts on that question. Here鈥檚 what he had to say.
鈥搁颈肠办
Rick: Just as a point of interest, how much masking is currently going on in schools?
Nat: Thirty-nine percent of students were required to mask under district mandates in our latest wave of data, which is from the week of February 28, 2022. That percentage is dropping and has already come down from 50 percent in the last week of January. I would expect that downturn to accelerate in the coming weeks as the last remaining state mandates expire.
Rick: Given that so many states are dropping their mandates, why is this tool relevant?
Nat: We know that masking is a bellwether for how localities are responding to COVID and how far from 鈥渘ormal鈥� they may be. In many places, I think masking provides concrete data that can be a proxy for COVID鈥檚 influence on people鈥檚 lived experiences. Even if through a glass darkly, that鈥檚 a useful indicator. Plus, this tool allows everyone to watch the progression of mask requirements loosening in schools. And more specifically, underneath the tracking you see on the website is a database that provides a whole additional layer of utility. With weekly data on mask mandates in over 8,000 school districts, researchers and decisionmakers at all layers of government can get to questions addressing why mandates fell when they did and how influential they were for managing pandemic risk. That large-scale, high-quality data can help us understand when masks make sense if pandemic risks surge in the future.
Rick: What might surprise people about what you found?
Nat: Media coverage on masks could easily lead readers to believe mask mandates have been up and down over time, but mandates have actually been pretty stable over this school year鈥攔emarkably so compared to the volatility of cases through the omicron surge. About 52 percent of students were under mask mandates at the start of the year, holding at 50 percent through the end of January. Over the following four weeks, we saw a drop of 11 percentage points, and that trend will only gain speed. Additionally, the contrast with current masking requirements and the CDC鈥檚 newly鈥攁nd massively鈥攃hanged masking guidance will surprise many. On February 28, the CDC changed guidance from 100 percent of students and teachers should mask in schools to only those in counties with 鈥淗igh Community Risk.鈥� Well, the current mask requirements we鈥檙e seeing don鈥檛 align with that guidance. For instance, 47 percent of students in low-risk counties are under mask mandates, while 66 percent of students in high-risk counties attend mask-optional districts. That pattern is upside down of what we would expect.
Rick: Has the recent wave of governors rescinding mandatory masking policies had a big impact?
Nat: Gubernatorial orders are very influential. That鈥檚 true in the case of states like Florida or Texas, which implemented bans on mask requirements in schools that were pretty clearly determinative. It鈥檚 also true in those that had state mandates. For instance, in Delaware, Governor Carney ended the state鈥檚 school mask mandate early and abruptly, effective March 1鈥攊n large part because of a court ruling. Once districts could make their own determination, all of them went mask-optional. Delaware students went from 100 percent attending schools with masking requirements one week to none being subject to mask requirements the next. I don鈥檛 think we will see as absolute a shift in our next update after New York, California, Oregon, and Washington鈥檚 mask mandates expire, but I鈥檓 predicting the trend will show the stark distance between local decisions and state requirements.
Rick: What other factors have played a role in explaining mask mandates?
Nat: There are other district characteristics that affect this, though I cannot say whether they were influences or mere correlations. District size was influential. Small districts, with just three to five schools, had consistently higher mask-optional percentages than did larger districts. Size and urbanicity were linked to more required masking, with rural districts much less likely to have had mask mandates all year. Meanwhile, high-minority districts and districts with historically-average achievement were more likely to have had mask mandates. I don鈥檛 see a case for causality there, but I do think that if masking inhibits effective schooling, or if it reflects a marked distance from normalcy, those patterns shed light on how the pandemic response may hamper recovery for groups hit hardest last year.
Rick: What, if any, is the connection between school closures and masking requirements?
Nat: It鈥檚 very strong. The website displays masking by districts depending on how much in-person instruction they offered last year鈥攂ased on the analysis in our Return to Learn Tracker. In 鈥淗igh In-person鈥� districts, 16 percent of students were under mask requirements last week; 62 percent were subject to mask requirements in 鈥淟ow In-person鈥� districts.
Rick: What鈥檚 one big thing we can learn from the masking data?
Nat: Perhaps it鈥檚 not surprising, but the big lesson that is so clear from comparing masking across types of districts is how mask mandates were linked to pretty polarized local cultural responses to COVID. Last year, when we tracked which districts operated in person or remotely, the correlation with presidential votes and vaccine hesitancy was glaring. Early pandemic local masking practices of the general public, as measured by in the summer of 2020, predicted remote instruction across the whole school year. Well, we see the same pattern between that 2020 masking survey data, collected two years ago, and today鈥檚 school mask policies. Put simply, how much localities were masking two years ago is a very strong predictor of school mask policies today, despite the monumental changes made by vaccines and the roller coaster of case rates. That suggests that local COVID culture鈥攖hat is, the local 鈥渃ommon sense鈥� regarding pandemic response鈥攊s driving current masking requirements more than anything else.
This interview has been edited and condensed for clarity.
