Masks work, even for kids. But sometimes science isn’t enough.

Masks are the most visible reminders that we are still in a pandemic, and the debate over their use and other public health measures has been present since March 2020, largely along partisan lines.

But as nearly a dozen Democratic governors — including in New York and New Jersey — announce an easing of mask mandates for schools and indoor business, it signals an overwhelming desire across the political spectrum for normalcy, regardless of the continued spread of the coronavirus.

Until last week, the removal of public health measures has essentially been happening in areas with Republican leadership, and public polling and academic studies have shown conservatives were much less likely to engage in social distancing, wear masks or support vaccination. Researchers who study partisanship, public health and human behavior say a new coalition is forming. The camp that never wanted COVID measures is being joined by people who’ve followed the rules for two years. They’re vaccinated; they’re boosted and they’re ready to return to normal.

“Those two camps coming together who are normally opposed, and now they are in greater alignment,” said Dr. Jay Van Bavel, a social psychologist and neuroscientist at New York University who studies how partisanship influences the way people think. “What you’re going to get is a bigger group of people who are not left or right. They could be all over the political spectrum — who have been boosted and done these all things for two years. They want a slow phase-in, or a quick phase-in, to normality.”

The Democratic governors have cited a rapid decline in COVID-19 cases as the rationale behind their rollbacks. But this pandemic has repeatedly had moments where the situation can be improving but still not great. New York state is averaging about 150 coronavirus deaths every day — a level three times higher than where things stood when Gov. Kathy Hochul issued a mask-or-vaccination mandate on December 10. Likewise, the number of people currently hospitalized with COVID-19 as of Friday (3,883) is just as intense as it was back then (3,603).

These leaders haven’t questioned the effectiveness of face coverings, but their recent decisions raise doubts about whether they’re fully following the science or looking for a quick way to appease the collective desire to get rid of masks.

It’s a question that’s long been at the center of the mask debate. Most studies show masks are protective for both adults and youngsters — without impairing their ability to breathe or read emotional cues. One study showed vocabulary memory was more accurate when kids were masked. But some of the loudest advocates desperately want to rollback mask mandates for schools — so much so that some claim face coverings don’t provide protection for young faces, that the evidence is “inconclusive,” or that masks are developmentally harmful.

Over the past month, media publications ranging from the Wall Street Journal to The Atlantic to USA Today to The New York Times have published opinion editorials making such claims — often authored by or citing the same circle of health experts. These editorials tend to omit the full body of research that supports mask usage, and instead cite a small collection of studies to fit their points of view.

The USA Today editorial, for example, stated “well-controlled real-world studies consistently fail to find a convincing benefit” for mask wearing, while linking to a story published by New York Magazine in August titled, “The Science of Masking Kids at School Remains Uncertain.”

In the New York Magazine article, journalist David Zweig questions the merits of a single study published in May 2021 by the Centers for Disease Control and Prevention that found COVID rates dropped in schools that required teachers and staff to wear masks. He also critiques a mask requirement for everyone older than 2, which the health agency recommended a few weeks later on July 9th.

But well before his story was published, and even before the CDC explained its reasoning behind the mask recommendation, a Facebook account with Zweig’s name was posting anti-mask opinions in a private group for a parenting community based in Hastings-on-Hudson, New York.

One comment by Zweig, posted on June 4th, 2021, alleges that New York schools in his area would keep mask mandates regardless of health guidance. In another from a year earlier, Zweig expresses support for an op-ed written in The Hill by Scott Atlas of the Hoover Institution, a controversial adviser to the Trump Administration who is regarded as a purveyor of COVID-19 misinformation regarding children that consistently contradicted public health guidance such as wearing masks.

The social media posts were shared by a concerned tipster with access to the Facebook group who requested anonymity due to fears of retaliation in their parenting community. In response to questions sent to New York Magazine, Zweig did not deny being the author of the Facebook comments, but defended his reporting.

“It is demonstrably true that many school districts have implemented measures that were never required or continued to implement measures after requirements were removed,” Zweig emailed to Gothamist. He also said that Atlas presumably made a point that he agreed with or thought interesting but it didn’t mean that he always agreed with the Trump adviser. New York Magazine did not indicate if it was aware of the Facebook posts prior to publication.

Like all adults navigating the pandemic, journalists likely have opinions on how health measures, like masking, affect their lives. But Zweig’s posts on a topic he is reporting on speaks to a larger trend experts see when science is part of a social or political debate.

“They will cherry pick studies that have smaller effects and are less conclusive rather than looking at the totality of the literature,” Van Bavel said. “When we’ve done research on climate change, you find, at least in the U.S. and a few other countries, it’s highly polarized. And when it’s polarized and people connect it to their identity …then it’s very hard to get through to those people.”

This cherry-picking is gaining momentum at a time when COVID-19 vaccination rates remain low among children. Only one in three New York kids ages 5-11 are fully vaccinated — which is only slightly higher than the national rate. Children, meanwhile, are experiencing more fatalities, according to CDC data. In 2021, COVID-19 took the lives of 539 children in the U.S. — most of which happened in the second half of the year.

The past two years of COVID-19 have killed more kids nationally than the total number of pediatric flu deaths recorded over the five years before the coronavirus emerged, when society was fully open.

As mask mandates end in the coming weeks, the omicron case surge could continue to decline for the overwhelming majority of people — while also still posing a danger to unvaccinated groups or at-risk communities whose immunities tend to fade faster. So in case the debate heats up again, here is a look at the evidence behind masks and the science behind partisanship.

Zweig’s New York Magazine article focused mostly on a CDC study conducted in Georgia toward the end of 2020. It looked at about 160 elementary schools with a combined 91,800 students.

The story hops back-and-forth between alleging flaws in the study and the CDC’s mask guidance, suggesting that the public health agency did not have “conclusive” evidence to support its July 9th recommendation that anyone older than 2 should mask indoors.

“Many people may surmise that even if there’s not conclusive evidence, it’s still likely that masking kids in schools helps,” Zweig wrote in the August 20th story in New York Magazine

When the CDC released its student-mask guidance in question on July 9th, it referenced a scientific brief containing more than 90 studies about masks, not just the one on Georgia. The mask section alone cited more than half a dozen reports.

Studies back then had already shown that masks could help impede superspreader events in places such as schools. One investigation in Israel revealed an uptick in cases at a school immediately after masking was rolled back for students. A separate episode involving Georgia schools linked student masklessness to COVID clusters.

“The Georgia study that was the focus of my article was, according to the CDC, the first of its kind to compare COVID-19 incidence in schools with certain mitigation measures — including mandated masking — to other schools without those measures,” Zweig told Gothamist. “This made it uniquely valuable relative to the other studies, which lacked the same comparative element or weren’t done in schools.”

Masks had also been shown to reduce airborne respiratory particles from entering and exiting people’s faces. At least one early investigation, published in April 2020, showed masks could block viral particles exhaled by people older than 11.

“Masking does physically block these aerosols,” Dr. Jennifer Dowd, an associate professor of demography and population health at the University of Oxford, told Gothamist in August 2021, after Zweig’s article was published.

The body of work supporting masks has only grown since then.

Massachusetts school districts found coronavirus attack rates were 10 times higher with unmasked interactions versus masked ones, according to a preliminary report released in September. A nationwide survey of more than 6,600 childcare professionals, published in late January, reported that child masking was associated with a 13% reduction in facility closures.

And a California study released on February 4th showed that mask usage with indoor settings, including at schools, reduced the odds of the positive COVID result by 44%, on average. Higher-quality masks — such as N95s and KN95 — lowered the chances by 83%.

When doubters try to downplay the value of face coverings, their arguments often usurp scientific language. One common complaint centers on the dearth of clinical trials on mask usage.

Two years into the pandemic, most people have likely heard of clinical trials for vaccines or other drugs. Such experiments are considered the gold standard of medical evidence because they attempt to remove internal biases that can skew results. Subjects are picked at random to receive either a treatment or placebo — and researchers are unaware of whom received what.

But some health interventions can’t be studied with clinical trials because it would be unethical to deny people something that could save their lives.

“We’re not going to do a randomized control experiment of whether a parachute saves your life when you jump out of an airplane,” Dr. Lawrence C. Kleinman, a pediatrician at Rutgers Robert Wood Johnson Medical School, told Gothamist in August ahead of the school year.

He said in order to do a randomized control trial on an intervention, one needs something called equipoise — that is two options for which there is truly no evidence that one is better than the other.

A Science Magazine report published in December is arguably the closest scientists have come. The study involved a randomized trial of approximately 342,000 adults spread across 600 villages in Bangladesh. All of the communities were offered masks to avoid the ethical dilemma of equipoise, but a subset were also given educational outreach or polite reminders from paid locals called mask promoters.

“So let’s say you see somebody on the street not wearing masks. We’d go up and say, ‘Look, we distributed masks for free to every household. We know you received one. Can you please put it on?’” Dr. Mushfiq Mobarak, an economics professor at Yale University who led the study, told Gothamist this week.

In villages without reinforcement, only about 13% of people wore face coverings. Mask wearing rates rose to 42% in places with the extra nudges, which corresponded with an 11% drop in COVID rates. Mobarak said the masking effect would likely be larger in other parts of the world.

“Life in rural Bangladesh is generally lived outdoors, and we now know due to better ventilation, et cetera that this is a relatively safe setting for COVID,” he said. “In places that are colder, places where people work indoors a lot more as opposed to farming out in the fields, you might think that masks are actually even more valuable than what our study shows.”

No, because inconclusive means no definitive result was found. Most mask studies may involve forensic investigations or retrospective looks at what happened during a superspreader event, but they still come with takeaways. This is another example of how scientific language and data are being misused to suit the opinion that face coverings lack worth.

Zweig’s piece in New York Magazine, for example, emphasized the fact that the Georgia study found a trend where COVID-19 rates were 37% in schools where teachers and staff used masks. The pattern was highlighted at the very top of the study and described as being statistically significant, a benchmark of supportive evidence.

Yet a similar trend for student masking — a 21% reduction in case rates — was not statistically significant and was only mentioned toward the end of the study. Zweig’s piece accuses the study of “file drawering” — or trying to bury a statistically flawed result. Zweig was arguing that reducing COVID cases by one in five could not be considered dramatic change without statistical significance.

Such a take exemplifies a basic misconception about what statistical significance is — and it’s a misunderstanding that statisticians have been trying to correct in mainstream media for decades.

Simply put, the absence of statistical significance for a trend does not mean a negative result was found. Statisticians often describe this rule as “don’t confuse no evidence of a difference” with “evidence of no difference” — but it essentially means a hypothesis is innocent until proven guilty.

“[No statistical significance] says something about the quality of the experiment or the thought experiment that the scientists performed,” said Dr. Jonathan Auerbach, a statistics professor at George Mason University who also served as a science policy fellow at the American Statistical Association. “But a low-quality experiment can produce non-significant results even when a treatment or a prevention strategy is very effective.”

No medical intervention is perfect. The COVID-19 vaccines started protecting against severe disease and the possibility of infection, but these powers wane over time and against new variants, necessitating booster shots. Masks can reduce the chances of airborne transmission, but they’re not foolproof. Remember the idea of the swiss cheese model of pandemic defense — the concept that masks and vaccines are single slices of the mitigation measures recommended to lower COVID-19 risk?

Auerbach agreed with Zweig that the study’s topline summary should have included the non-significant findings, but “that doesn’t necessarily mean that those results prove that the [mask] treatment is ineffective,” he said.

Omitting relevant studies from the New York Magazine story didn’t stop the article from going viral on social media. Since then the general public, news pundits, talkshow hosts and even education reporters have shared it when they want to knock down the idea that masking kids reduces coronavirus rates.

And Zweig has continued to publish similar articles.

Zweig wrote a story in The Atlantic wherein he again described the CDC’s scientific case for mask wearing as “flawed” — while mostly referring to just one study. This story also cherry picked information, such as when it said “The World Health Organization, for example, does not recommend masks for children under age 6.” But the WHO also says children ages 6-11 should wear masks in areas with widespread transmission and that face coverings should be mandated for everyone 12 and up.

Van Bavel from NYU said the situation is similar to decades of conversation around climate change. From a scientific perspective, the evidence around the human contributions to the warming of the planet through carbon pollution is firm and settled. Most Americans believe global warming is happening and are noticing an uptick in extreme weather events, according to the Yale Program on Climate Change Communication and Pew Research. Even a small but growing portion of conservatives are starting to believe.

But there are also still hundreds of thousands of Americans opposed to the concept of human-caused climate change, and research shows their opinions typically don’t change without experiencing an extreme weather event.

“When we asked people why have you changed your mind about global warming? The single most common answer was that I’ve experienced the impacts,” said Anthony Leiserowitz, director of the Yale Program on Climate Change Communication.

Some people will always see what they want from COVID data and studies. In his New York Magazine piece, Zweig writes “Out of more than 600,000 American deaths attributed to COVID-19, 361 were for kids and adolescents under age 18. In the much-shorter 2018-2019 flu season, there were 477 pediatric deaths.” But in reality, the article is citing an “estimated” projection for the flu mortality that is based on hospitalizations, rather than the actual count of flu pediatric deaths.

In response, Zweig told Gothamist that “the CDC provides these estimates for pediatric flu deaths specifically because it believes the ‘hard’ counts of pediatric flu deaths are inaccurate.” That’s true because some unknown number of flu cases are missed every year — but the same can be said for COVID-19. Scientists predict 30% of coronavirus cases are asymptomatic and missed. Regardless, the New York Magazine article is making an apple-to-oranges comparison — between a flu estimate and a COVID count.

Van Bavel from NYU said some emerging research from Syracuse University showing that there are people who didn’t want any type of COVID measure from the very beginning of the pandemic. According to the manuscript, which is under review with a major academic journal, members of this hardcore group have remained the same.

“You’ve seen this in political discourse and since the very beginning of the pandemic in the United States,” Van Bavel said. “There’s people who didn’t want lockdowns from the beginning. They didn’t want any impingement on their freedoms. They thought the virus was a hoax or just the flu.”

At the same time, many politicians made a deal with the public last year: Get vaccinated, and people can take off your mask. But this bargain pitted vaccines against masks in the mind of the public.

“There is a reasonable person’s argument, which is once we got the vaccine and we’re boosted, there is actually less need for us to wear a mask because we’ve dramatically minimized our probability of hospitalization or mortality,” Van Bavel said. “The question is when?”