The Washington PostDemocracy Dies in Darkness

The mask mandates were inside of us all along

A mask lies abandoned at Liberty Hill High School in Texas on Jan. 20. The school was temporarily closed because of staff shortages. (Matthew Busch/Bloomberg News)

As I was writing about the sudden push by Democratic governors to rescind mask mandates last week, I noticed something interesting.

My overall point was that new case totals in the states that were scaling back mask rules had declined precipitously after the emergence of the omicron variant in December — but that the governors appeared to be responding more to political expediency than any sort of planned response to the data. Pulling the state-level case totals, though, it became obvious that there wasn’t any significant difference between the places that had instituted or maintained mask mandates and those that hadn’t.

Sign up for How To Read This Chart, a weekly data newsletter from Philip Bump

That’s obvious if we look at every state. Those that implemented new mandates to respond to omicron or had ones in place (using data from AARP and Ballotpedia) are indicated in orange. Those that didn’t are in black.

There is not any detectable pattern. (Notice that these are population-adjusted, so the spike in New York, for example, is not simply because it is more populous.)

In fact, if we cluster those states with mandates (about 28 percent of the population) with those that didn’t have any, we see that there’s not much difference in the rise and fall from omicron — and that the mandate states had slightly higher per capita case totals.

There are a lot of ways in which this could be explained. It is possible, for example, that mask mandates were put in place in states that were already seeing big surges and the mandates tamped down on how bad they would be. Or that the mandates came too late to prevent large surges. States with mandates were also more heavily vaccinated; perhaps confidence in the efficacy of the vaccine to prevent infection led to fewer precautions. The mishmash of rules and implementation points and other factors make such theories hard to immediately evaluate.

But there’s probably a simpler explanation. Consider polling from YouGov conducted for the Economist just before the omicron surge, at its peak and last week, after the peak faded.

The differences are subtle. Before the surge, slightly more than half of the respondents said they wore a mask always or most of the time. Now, 6 in 10 do. The percentage of respondents who say they never wear a mask went from 22 percent in November to 16 percent in January to 19 percent now — shifts that are not statistically significant.

Perhaps there were people who lived in mandate states (again, just over a quarter of the country) who moved from “never” to “most of the time” as a result of those mandates. Or perhaps this captures a general pattern in how people approach mask-wearing, that some people keep wearing masks regardless of mandate and some don’t. That would suggest that infection patterns would be similar regardless of mandates.

Of course, there’s also the partisan difference. In YouGov’s most recent poll, Democrats were twice as likely as Republicans to report wearing masks. But more Democratic states (as measured by 2020 vote) reported higher peak cases during the omicron surge. (Democratic states that supported President Biden more narrowly had a lower peak than any other group.)

The answer here isn’t clear. Both these data and observed reality, though, suggest that it’s probably fair to assume that individual choices had more effect during the surge than state-level mandates.

Another reason that rescinding mandates might have been better-positioned by Democratic governors as something more substantive than a response to the unpopularity of the rules.