Masks and counterfactuals
Why I don't post many “masks don’t work” takes
I don’t run many mask takes because most of them respond to a misunderstanding that is itself a bad take. It’s hard to untangle this in tweets, so I offer an explanation here.
First, let’s look at some exemplary anti-mask tweets of the type I’ve avoided so far:
Why “masks don’t work” is a bad take
All of these are responding to the incorrect assumption that masks would be absolutely effective—that masks by themselves would completely stop transmission of the disease.
But this isn’t why we wear masks. They weren’t supposed to be a silver bullet.
In fact, from the very beginning cloth masks were recognized to be an imperfect improvisation. We didn’t want to endanger PPE for healthcare workers and figured “cloth is better than nothing.” Nine months later, it’s incredibly frustrating that our nation has never made more efficacious and credibly-sourced masks widely available. Nine months later, most of us are stuck with random cloth and dodgy imports filtered through the grey market.
Cloth masks were never supposed to end the pandemic by themselves, just help slow it in conjunction with other mitigation. Masks—even the improvised cloth masks that predominate in America—probably do reduce the spread of the virus, but they cannot suppress it entirely. They appear effective, but not 100% efficacious.
The bad takes go astray because they pretend mask were supposed to stop the disease on their own. They point to the admitted spread of disease and say “aha! masks don’t work!” But traffic deaths don’t prove seat belts don’t work. Likewise for masks.
“But,” say the antimaskers, “how do you explain cases rising in places that have high mask usage, like Los Angeles?” The answer is: the same way I (and anti-maskers) explain South Dakota (with relatively low mask compliance) getting much higher rates of infection and death compared to other rural states starting in September.
COVID-19 is a strongly seasonal disease.
Los Angeles in January is worse off than Los Angeles in September, true, but this isn’t surprising given seasonal trends. The California example only shows that masks by themselves do not reduce R0 below 1.0 (…in LA…in December), but we can’t conclude they do nothing.
In order to determine that masks “don’t work,” you need to compare what would have happened in their absence. You need to compare the counterfactual. Unfortunately, we can’t access the alternate reality where California didn’t wear masks at all, but we can guess that low mask compliance might be one of the reasons more people per capita died in rural South Dakota than more urban Illinois. South Dakota’s rural nature and later onset of the pandemic should have given it a couple of advantages.
Perhaps there are other confounding variables that makes my South Dakota comparison unfounded, but simply pointing to infections does not prove masks “don’t work” at the margins.
Why I rarely run “masks don’t work” takes
While the spread of these bad takes is based on a misunderstanding of the pro-mask position, I can’t call it a strawman. A lot of people pretend that universal mask wearing would halt the spread of the virus.
As an example, some of my followers were upset I featured this take yesterday:
I ran this primarily for the reasons Zeynep Tufekci laid out. It’s a smug message that excoriates people who in all likelihood have been wearing masks. This isn’t a good way to encourage civic-minded cooperation. It’s a good way to recruit anti-maskers who read the message in the graphic and respond “no, it’s because government tyrants and smug white-collar internet assholes want to blame their problems on people like me.”
The graphic also perpetuates the mistaken idea that if there were 100% mask public mask compliance the pandemic would end. And people who internalize this idea will eventually find a skeptic’s chart of Los Angeles infections and mandates over time and conclude pro-maskers are full of it.
I’m always wary of running takes that reinforce misguided worldviews, so I will continue to run only the most egregious “masks don’t work” takes for now.
I hope that folks who take COVID seriously present a more nuanced case in favor of masks; they’re a low-cost way to reduce the spread of the disease and people should wear them because it will occasionally help to prevent some deaths. People will still get sick, we should continue to be careful of indoor environments, especially small spaces, and behaviors that produces lots of droplets like shouting and singing. But I believe people can be talked small helpful acts even if they’re skeptical.
EDITS Jan. 14: Added Neil Clark take, expanded some points and links.