In response to on-going concerns about COVID-19, in July 2021, the Centers for Disease Control and Prevention (CDC) recommended universal masking in kindergarten through grade 12 schools during most of the 2021–2022 school year for students, teachers and visitors, regardless of individuals’ vaccination status or community transmission rates (1, 2). At some point during the 2021–2022 school year, 18 states and the District of Columbia implemented mask mandates for public (and in some cases, private) schools in their jurisdiction (3–5), while other states did not enact mask mandates or expressly prohibited them.

In February 2022, however, the CDC updated its guidance to recommend masking in schools only in the context of high community transmission or strained healthcare system capacity (3, 6).

Following this change in CDC masking guidance, many of the 19 states (including the District of Columbia) rescinded previously implemented mask mandates in schools (3). In response to state-level policy changes, some school districts lifted mandates while others chose to retain them. A recent study in Massachusetts, the only study so far investigating factors associated with school district decisions on mask mandates, found that school districts that lifted mask mandates experienced greater incidence of COVID-19 among students and staff compared to those that did not (4). Moreover, districts that continued their mask mandates after the statewide mandate was lifted were more likely to serve students from minoritized racial and ethnic groups and students in poverty (4). To date, however, factors associated with mandates in districts across the US have not been evaluated.

The impact of COVID-19 has varied substantially across communities in the United States. In light of findings from the Massachusetts study, it is possible that, nationwide, communities that bore a greater burden or COVID-19 related morbidity and mortality risk adopted were more likely to adopt more stringent masking policies; communities of color and lower socioeconomic status have been disproportionately impacted (7, 8). Urban and rural areas have also been impacted differently; while cases overall were initially concentrated in urban areas, intensity shifted between urban and rural areas over time (9).

Prior research at the national level suggests that individual characteristics like political affiliation (10), race (10, 11), gender (11), urbanicity (10), and income (10) are associated with the acceptability of masking as a public health strategy and adherence to masking requirements. Limited research has also identified factors associated with the presence of statewide mask mandates, notably political party affiliation of the state’s governor; states with Republican governors have been less likely to implement statewide mask mandates (12, 13).

The objective of this study was to characterize student mask policies for all 50 US states and the District of Columbia (DC) and a sample of United States school districts at four points during the 2021–2022 school year: September 2021, November 2021, January 2022, and March 2022. We then sought to determine whether urbanicity, demographics of district students, political affiliation of the state governor, or community viral transmission levels predicted the presence of a school mask mandate at the school district level at any time point. Finally, we examined whether the factors that predicted a mask mandate at any time point differed from those that predicted retaining a mask mandate at all four time points, including after the CDC’s masking guidance was revised. We hypothesized that school districts in states with Democratic governors, those in urban areas, those with a higher proportion of students from minoritized racial/ethnic groups or economically disadvantaged students, and higher county-level COVID-19 transmission levels would be more likely to have mask mandates. We also hypothesized that district demographic composition would be more strongly associated with the likelihood of the most conservative masking approach (i.e., requiring student masking at all four time points) as compared to the likelihood of having a mask mandate at any time point.

In this study, we aimed to characterize student mask policies for US states and a sample of United States school districts at four points during the 2021–2022 school year and to determine whether urbanicity, demographics of district students, political affiliation of the state governor, or community viral transmission levels predicted the presence of a school mask mandate. We found that having a school mask mandate was related, most robustly, to the political affiliation of the state governor and urbanicity of the district, although these relationships changed over time.

Our findings echo prior literature noting the link between political affiliation and urbanicity to mask wearing, and the presence of statewide mitigation policies (10, 12, 13). However, this study extends prior findings to the school district setting where political affiliation and urbanicity similarly continued to influence mask policy at the school level.

While individual factors (namely, governor political affiliation, and urbanicity) were associated with the presence of district level student mask policy in September 2021, November 2021, and January 2022, there was no association between any of the factors we examined and the presence of a mask mandate in March 2022. By March 2022, all but 20 districts were either recommending mask wearing or making it optional. Of the remaining 20 districts that did not have an optional/recommended policy in March, the 12 that were in states with Democratic governors continued with their mask mandates. This could be due to lower COVID-19 case rates at this time, the February shift in CDC guidance, rising student vaccination coverage, or collective exhaustion with COVID-19 mitigation that shaped public and political appetites to continue masking requirements (24).

Generally, districts that served a larger proportion of students from minoritized racial/ethnic groups were more likely to have a mask mandate at any timepoint (upper third quartile of districts) and all (upper third and fourth quartiles of districts) timepoints. This is consistent with literature that describes high rates of mask wearing in communities of color overall (11) and could reflect greater demand for COVID-19 mitigation in communities disproportionately impacted by COVID-19 (7). Consistent with this hypothesis, Cowger et al. (4) found that schools in Massachusetts that sustained mask policies beyond when they were required tended to have greater proportions of Black and Latino students and staff.

We also found that, as the proportion of students with economic disadvantage increased, the odds of having a student mask mandate at all time points decreased. These results differ from Cowger et al. (4), who observed greater persistence of mask mandates in districts with a higher proportion of low-income students, but align with those of Kahane (10), who found that greater county-level median household income was associated with greater masking behavior. Our results should be interpreted in the context of a sample with relatively high average (64.8%) economic disadvantage based on our sampling strategy; further research should clarify the role of economic disadvantage in masking behavior and policy.

Notably, relative COVID-19 burden was not a significant predictor of a district mask mandate at each, any, or all of the four timepoints included in the analysis. This is salient as evidence indicates that school masking is an effective way to mitigate COVID-19 spread (25–27). However, what might be considered a directly relevant, public heath factor (COVID-19 prevalence) did not seem to be important in district mask policy decision making; district mask policies seem to be driven by factors such as political party, urbanicity, and local racial/ethnic composition, which in turn have tangible public health consequences for communities.

There are some limitations in our analysis. First, our school district level dataset was limited to data from 20 states, and purposive sampling was used within each state to select included districts. Relative county-level COVID-19 burden (high/medium/low by month) was used as a proxy for district COVID-19 incidence; however, in some places, school districts and counties are not interchangeable. In addition, our mask policy data are based on publicly available information from state and district websites and may not always reflect the most current policies that were being enforced.

Although mask wearing is an evidence-based intervention for reducing COVID-19 transmission, key questions remain about the degree to which school mask mandates impact the health of students, staff, and the community. Another important question is whether mandates negatively impact learning, socialization, or well-being and, if so, for which students. Our results raise ethical concerns about what factors should influence school health policies during a pandemic. Political affiliation of a state governor and district urbanicity may reflect community values about balancing protecting health with parental and student freedoms and other student interests, but they may also underplay or overplay the public health benefit, and the benefit to individual students and families, of mask mandates.

We found that various social, economic, and political factors predicted school mask policies for students, and that these associations varied over the course of the 2021–2022 school year. Our findings underscore the importance of learning from the experience of this pandemic to provide the best possible evidence of the benefits and harms of mask mandates. Although the next pandemic may have different epidemiologic and clinical features relevant to mask policy, a more robust evidence base will inform the development of more centralized, consistent guidance that may help narrow differences in mask policy by politics and geography. Additionally, public health officials should be mindful to deliver guidance in a way that anticipates and/or minimizes political reactions and enhances trust when attempting to influence school policy.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

SJ, AA, KB, RF, MK, AN, AR, AT, and MC were instrumental to the creation of the dataset for analysis and the conception of the study. XG and LK led data analysis for the manuscript. LK led drafting and synthesis of results for the manuscript. All authors contributed to the article and approved the submitted version.

LK, SJ, AA, KB, RF, XG, MK, AN, AR, AT, and MC disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Johns Hopkins eSchool+ Initiative.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.