This summer, Indiana's Governor Eric Holcomb decided not to reinstate mask mandates in Indiana schools. Despite recommendations from the U.S. Centers for Disease Control and Prevention and the American Academy of Pediatrics that universal mask-wearing be used to prevent the spread of COVID-19, Holcomb left decisions about mask mandates to local school boards. Just 38 of 290 public school corporations—representing 29 percent of K-12 enrollment in public non-charter schools—started the year with mask mandates. Since starting the school year, many other school corporations have added mask mandates.
Indiana public schools that started the school year with a face mask mandate are reporting fewer COVID-19 cases than those that did not enact mask mandates. While reported cases spiked in both groups in late August, the gap between mandate and non-mandate schools remained. A third group of school districts began the school year without a mask mandate but enacted one in the first few weeks of school. The overall infection rate in these school districts was higher at the beginning of the school year, but has since fallen below the rate at districts without a mandate.
Data sources: Indiana MPH, Indiana DOE, David Waldron
School mask mandates and community spread
Not only did school districts with mask mandates have lower COVID-19 infection rates in schools, they also have lower rates of community spread than areas without school mask mandates. An analysis by ZIP code shows a similar pattern to the school district analysis, with cases increasing in both groups, but staying lower in areas with mask mandates.
Data sources: Indiana MPH, Indiana DOE, David Waldron
This ZIP code-level analysis also allows inclusion of other data, such as vaccination rates. Below, the areas are categorized into three roughly equal groups by the percent of population fully vaccinated in early September:
- Low vax (40% or less)
- Mid vax (40% to 50%)
- High vax (over 50%)
The story doesn't change. In each vaccination category, areas with mask mandates still have lower infection rates than those without mandates.
Data sources: Indiana MPH, Indiana DOE, David Waldron
What drives mask mandate adoption?
School policies have changed rapidly in the first few weeks of school. Just 29 percent students in Indiana public school corporations began under a mask mandate. But as of September 18th, 65 percent was under a mask mandate. Heated debates about mask mandate policies broke out across the state at school board meetings in August.
Politics appears to be a major factor in whether a school started with or eventually added a mask mandate. In school districts that voted for Hillary Clinton in 2016 (detailed 2020 data is not yet available), 83 percent of students started with a mandate, rising to 98 percent weeks later. The percentage in districts where Donald Trump won at least 70 percent of the vote started at 2 percent and has risen to just 29 percent. The prevalence of mask mandates where Trump received between 50 percent and 70 percent of the vote saw a dramatic increase from 22 percent early in the school year, to 77 percent by September 18th.
This association between politics and public health measures is unfortunate, as many people turn to disinformation instead of expertise. But it also makes it difficult to measure the impact of policies like school mask mandates. Generally, it suggests that school districts without mask mandates might differ from those with mandates in many ways that would also influence the spread of COVID-19. Of districts that started without mask mandates, 81 percent are rural, while only 19 percent of districts that opened with mandates are rural. Rates of educational attainment also differ between the two groups: the rate of college graduates in mandate districts is 31%, compared to 20 percent in districts with no mandate. This difference in educational attainment could reflect further differences in social structure and occupational exposure to infection.
In the chart below, it is apparent that school districts with mask mandates tend to have lower COVID-19 infection rates, but it is not clear that there is a difference between mandate and non-mandate districts that are similar politically. School districts that changed mask mandate policies after school started are excluded, since policy changes during the school year are more likely to be reacting to infection rates.
Data sources: Indiana MPH, Indiana DOE, VEST, David Waldron
In both analyses of school cases and community spread, the main barrier to making conclusions about the effectiveness of mask mandates is that areas with mask mandates are different in many ways from areas without them. Other recent studies of school mask mandates (one of public schools in two Arizona counties, and another nationwide county-level study) come to similar conclusions, but suffer from similar limitations. The best-quality evidence for the effectiveness of mask-wearing on COVID-19 spread comes from a randomized trial in Bangladesh, which showed that an increase in mask-wearing from 13 to 42 percent caused a 12 percent decrease in the prevalence of COVID-19 symptoms, and a 9 percent decrease in COVID-19 symptomatic seroprevalence.
Data Availability
Data used for this post is available for download here.
