We need better masks, and better approaches to using them.
Two weeks ago, I wrote about some reservations I have about a Cochrane review meta-analysis finding mask-wearing in the community probably “makes little or no difference” against respiratory infections. While the overwhelming majority of the studies included in the review weren’t about Covid — and the ones that were looked substantially more promising than the studies of masking outside a pandemic — it’s been widely taken as proof that mask mandates never worked against Covid.
I think the Cochrane review is flawed, for the many reasons I discussed in the original newsletter, but it attempted to answer a question that people are completely right to care deeply about. There will almost certainly be a next pandemic, and a good understanding of the costs and benefits of different public health interventions will be crucial to figuring out how to handle it.
Knowing whether masks have any benefits is one critical piece of that puzzle. Rather than being the last word on the subject, the Cochrane controversy should invite us to dig into the question instead of further polarizing us around overly simplistic answers to it.
What we talk about when we talk about masks
Jason Abaluck is a Yale behavioral economist and the co-author of a study of masking interventions in Bangladesh that provided the single largest randomized controlled trial we have that looks at the effects of encouraging people to wear masks during the Covid pandemic. His research found a significant reduction in Covid cases in the villages that encouraged masking.
When I talked with him this week, he emphasized that his finding — even if it’s eventually supported by further research — is only one piece of the puzzle when it comes to figuring out masks.
“There’s a couple distinct questions that are getting conflated here,” he told me. Here are some of them:
“If you wear a mask while there’s a pandemic, are you less likely to get sick?”
“If you wear a mask while you have a respiratory illness, are you less likely to infect other people?”
“If you make people less likely to get sick during a pandemic, does that have lasting benefits to them, or does it just delay an infection without significantly changing their long-term health outcomes? Does it reduce transmission enough to change the overall dynamics of the pandemic?”
“If you tell people to wear masks, will they actually wear masks correctly and reliably?”
“If you mandate that people wear masks, will they actually wear masks correctly and reliably?”
“What are the costs, to the median person and to a person who is unusually affected by wearing masks, of wearing masks?”
Better politics through better science
It should be obvious that in order to make public policy decisions about mask mandates, we would ideally know the answers to — or have good guesses for — all of these questions. Instead, we’re often stuck treating one question — do masks reduce respiratory illness? — as a proxy for the policy question of “should there be mask mandates during pandemics?”
The conversion of the question of whether masks reduce illness into a proxy for a hot-button political issue is bad for science and for politics. We’ll make the best policy decisions if we have real answers on each of these questions, and, after years of a pandemic, we should really be doing far more to get those answers.
Here’s what we know: It takes a lot to get people to wear masks. Information campaigns have a fairly limited effect. Mandates have a much larger one. “A general stylized fact in behavioral economics,” Abaluck told me, is that “it is very hard to get people to do anything differently than they were doing beforehand.”
Increases in mask-wearing do appear to reduce disease transmission during a pandemic, though it’s hard to extrapolate how large that increase would be if everyone wore masks.
The Bangladesh study looked at the effects on mask-wearing of various measures to encourage people to wear masks (they found a suite of interventions that increased mask-wearing by about 30 percent) and then looked at the effect of that higher masking rate on symptomatic Covid cases (they’re about 10 percent lower), which gives us a glimpse at the answer to some of these questions. But ideally it’d be accompanied by a lot more research than we have now.
Superior technology makes social problems easier to solve
There’s one set of interventions that I am excited about regardless of the state of existing research on masks, and that is better masks.
Let’s face it, the masks we all had to wear during Covid were frequently quite unpleasant. My glasses fogged up so I couldn’t see indoors; wearing masks for a long time came with ear pain; it was harder to understand people’s speech, especially for people who are hard of hearing; and some research indicates that masks may have inhibited the development of kids’ speech and social skills.
That trade-off may well have been worth it, and probably was worth it for many people at the peak of Covid infection risk and early in the pandemic, when delaying infections and preventing hospital overcrowding was vital. But one way to make masking a better public health intervention is to decrease its costs by figuring out better protective equipment that’s less unpleasant to wear and that filters out more pathogens.
There could and may well be a pandemic much deadlier than Covid, and if there is, some pretty strong measures would be justified to bring it to a halt before it sweeps the world. Good research into whether masks as worn by normal people under normal conditions prevent disease transmission is a big part of figuring out what our pandemic response should be.
But another part is developing better options. Right now, we’re stuck relitigating the mask wars because, despite Covid’s warning shot, our options should a pandemic emerge tomorrow are essentially no better than they were three years ago. While researchers figure out exactly how well masks work and exactly what that should mean, they should also be working on giving future policymakers better choices.
A version of this story was initially published in the Future Perfect newsletter. Sign up here to subscribe!