medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022.

The copyright holder for this preprint

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Impact of Lifting School Masking Requirements on Incidence of COVID-19 among Staff and
Students in Greater-Boston Area School Districts: A Difference-in-Differences Analysis

Tori L. Cowger1,2, Jaylen Clarke2, Eleanor J. Murray2,3, Sarimer M. Sánchez2, Mary T. Bassett1,
Bisola O. Ojikutu2,4, Natalia Linos1, Kathryn T. Hall2,4*

1. FXB Center for Health and Human Rights, Harvard T.H. Chan School of Public Health
2. Boston Public Health Commission
3. Department of Epidemiology, School of Public Health, Boston University
4. Brigham and Women’s Hospital, Harvard Medical School

*Corresponding Author:
Kathryn T. Hall
Deputy Executive Director, Boston Public Health Commission
1010 Massachusetts Ave. 6th Floor
Boston, MA 02118
[email protected]

Conflict of Interest Statement. All authors have completed the ICMJE uniform disclosure form at
www.icmje.org/coi_disclosure.pdf and declare no competing interests.

Funding Statement. No external funding was received for this study.

Page
NOTE: This preprint reports new research that has not been 1 of
certified 23 review and should not be used to guide clinical practice.
by peer
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

ABSTRACT

Background. In February 2022, following the rescinding of a Massachusetts statewide school masking
mandate, only two cities (Boston and neighboring Chelsea) out of 79 school districts in the greater-
Boston area, maintained masking requirements in K-12 schools. This provided an opportunity to
examine the impact of removing masking on COVID-19 case rates among students and staff in the
public-school setting.

Methods. We used difference-in-differences for staggered policy adoption to compare incidence of

COVID-19 cases among students and staff in greater-Boston area school districts that lifted masking
requirements to those that had not yet lifted masking requirements during the 2021-2022 school year.

Results. Before the statewide school masking policy was lifted, there was no statistically significant
difference in case rate trajectories between school districts. However, weekly and cumulative case
rates were significantly higher in students and staff in school districts that removed masking
requirements, compared to districts that had not yet lifted requirements. We estimate that lifting of
school masking requirements was associated with an additional 44.9 (95% CI: 32.6, 57.1) COVID-19
cases per 1,000 students and staff over the 15 weeks since the lifting of the statewide school masking
requirement, representing nearly 30% of all cases observed in schools during that time. School districts
that sustained masking requirements for longer periods tended to have older school buildings in poorer
condition, more crowded classrooms, higher proportion of low income and English learning students
and students with disabilities, and a higher proportion of Black and Latinx students and staff.

Conclusions. Masking is a relatively low-cost but effective intervention that can protect students and
staff from substantial illness and loss of in-person days in school. Despite compelling evidence that
masking significantly reduces the spread of SARS-CoV-2, political will and public adherence to masking
has waned. Our study confirms that universal masking requirements can benefit all students and staff,
and therefore represents an important strategy to mitigate the impacts of structural racism, ensure
health equity, and to avoid potential deepening of educational inequities.

Page 2 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

MANUSCRIPT

INTRODUCTION

The direct and indirect impacts of the COVID-19 pandemic on children, their families, and surrounding
communities have been substantial. Through February 2022, more children and adolescents in the U.S.
have been infected with SARS-CoV-2 than any other age group.1 Mass infection during the Omicron
wave was especially consequential for younger age groups. During December 2021 - February 2022,
approximately one-third of all children and adolescents in the U.S. were newly infected with SARS-
CoV-2, and peak hospital admissions for COVID-19 were at least 3 times as high as any other period
during the pandemic.1,2 While the risk of severe COVID-19 is markedly lower compared with adults,
children with COVID-19 are at risk for severe acute complications, including Multisystem Inflammatory
Syndrome (MIS-C) as well as persistent long-term sequelae (i.e., long COVID/post-COVID
conditions).3–18 Layered on these health impacts are substantial impacts to children’s social
environments. More than 140,000 children and adolescents in the U.S. were estimated to have lost a
parent or caregiver to COVID-19 through June 2021.19 In addition, COVID-19 has caused substantial
interruptions in school settings, including staffing shortages, school closures, and missed school days,
deepening existing educational inequities.20,21

Importantly, the impacts of COVID-19 have been disproportionately borne by groups made vulnerable
by historic and contemporary systems of oppression, including structural racism and settler
colonialism.22–25 Black, Indigenous, and Latinx children and adolescents were more likely to experience
severe COVID-19 outcomes, death of parents/caregivers, worsening mental health, and educational
disruptions, among other outcomes, compared to their white counterparts.3,19,21,26–29 These
manifestations of structural racism underscore the urgent need to prioritize health equity in COVID-19
policies and programs that impact children and adolescents.

Schools represent an important venue for policies and interventions that minimize COVID-19’s impacts
on students and staff. Even prior to the COVID-19 pandemic, schools were not uniformly health-
promoting environments, shaped by environmental racism and chronic underinvestment. Historic and
contemporary policies and practices including redlining, land theft, disinvestment and gentrification
have eroded tax bases and shaped the quality of public school infrastructure and associated
environmental hazards.23,30–35 These varying school conditions, including crowded classrooms,
exposure to toxins and pests, and poor air quality due to outdated or absent HVAC/ventilation systems

Page 3 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

have left school districts differentially-equipped to respond to COVID-19, with harms concentrated
among low income and Black, Latinx, and Indigenous communities.30,36,37

Alongside improved ventilation, vaccination, testing, and social supports to minimize the secondary
impacts of COVID-19, masking represents an important piece of a layered mitigation strategy in school
settings.38–41 A growing body of evidence supports the effectiveness of universal masking requirements
in reducing SARS-CoV-2 transmission both in community and school settings.42–48 During the
emergence of the Delta variant prior to the start of the 2021-2022 school year, CDC issued guidance
recommending “... universal indoor masking by all students, staff, teachers, and visitors to K-12
schools, regardless of vaccination status”.49 However, on February 25, 2022, CDC released updated
guidance that limited masking recommendations in public indoor settings (including K-12 schools) to
counties with high COVID-19 Community Levels – a CDC-defined metric largely determined by COVID-
19 hospitalizations.50 Following this updated guidance, many statewide policies requiring masks in both
community settings and schools were lifted, largely shifting policy decisions to the local level.51

Massachusetts (MA) was one of only 18 states plus DC with statewide school masking requirements at
some point during the 2021-2022 school year.52 The Massachusetts Department of Elementary and
Secondary Education (DESE) lifted the statewide school masking requirement on February 28, 2022,
joining a number of other states, from California to New Jersey, that similarly removed mask mandates
in schools around the same time.52 With statewide school masking orders no longer in place and newly
revised CDC guidance for public indoor settings, many MA school districts lifted masking requirements
immediately, several sustained masking requirements for several weeks following the rescindment of
the statewide order, and two districts – Boston and neighboring Chelsea Public Schools – maintained
masking requirements through June 2022. The staggered lifting of masking requirements across
Boston-area school districts presents a unique opportunity to evaluate the impact of school masking
policies, during a period of highly transmissible SARS-CoV-2 variants and a rapidly changing COVID-
19 policy environment.

The goal of this study was to examine the impact of staggered lifting of school masking requirements
on the incidence of COVID-19 among staff and students in MA school districts, and to describe
potential impacts of policy choices for health equity. Specifically, we aimed to (1) compare weekly
incidence of COVID-19 in school districts that lifted masking requirements to districts where masking
requirements had not yet been lifted; (2) estimate the difference in COVID-19 incidence among
students and staff attributable to lifting mask protections (i.e., excess risk and population attributable
fraction); and (3) compare school districts characteristics (e.g., staff/student sociodemographics, school
Page 4 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

building conditions, etc.) for districts that chose to sustain masking policies for longer periods of time to
those that lifted masking requirements earlier.

METHODS

Data Sources & Definitions

We used publicly-available, school district-level data on COVID-19 cases, enrollment, and staffing for
the 2021-2022 school year from DESE.53,54 Each Thursday, school districts are required to report
COVID-19 cases among students and staff for the prior 7 days (Thursday-Wednesday). In addition to
required case reporting throughout our study period, DESE offered state-sponsored testing programs
options, including pooled and symptomatic testing and option introduced in mid-January for state-
provided weekly take-home rapid antigen tests for schools that opt-in.55,56

For included school districts, we manually gathered data on dates of school and citywide masking
policies from school district websites and/or local news sources as available. In addition, we obtained
data on COVID-19 indicators in the surrounding communities (case rates, percent test positivity, testing
rate) from the Massachusetts Department of Public Health (DPH) used for covariate adjustment in
sensitivity analyses described below. Finally, for descriptive analyses, we extracted information on
school district characteristics, including demographic information for students and staff, enrollment of
DESE-defined selected populations (e.g., low-income students, English learners, students with
disabilities, etc.) from DESE,54 and information on school building condition and learning environment
from the 2016 School Survey (most recent available data) from the Massachusetts School Building
Authority (MSBA).57

Exposure and Outcomes

The primary exposure in this study was whether a school district maintained or lifted masking
requirements in each reporting week. Under the statewide school masking requirement in place through
February 28, 2022, all schools had masking requirements in place at the start of our study. A school
district was considered to have lifted their masking requirement if their policy was rescinded prior to the
first day of the reporting week (i.e., if the masking requirement was lifted part way through the reporting
week, that week was classified as having masking requirements sustained).

Page 5 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Our primary outcome of interest was weekly reported rates of COVID-19 among staff and students. We
reported rates as the number of COVID-19 cases per 1,000 staff and students (considered together),
and separately among staff and students.

Inclusion and Exclusion Criteria

For this analysis, we considered the n=79 school districts within the greater-Boston metro area
contained within the U.S. Census Bureau-defined Boston-Cambridge-Newton New England City and
Town Area (NECTA) division after excluding Charter and Vocational/Technical school districts (Figure
S1). Of these, we excluded n=7 school districts with unreliable or missing testing data for more than 5
weeks of the study period (Supplementary Appendix, Methods). Our final sample included n=72
school districts representing n=294,084 students and n=46,530 staff over the 40 calendar weeks of the
2021-2022 school year through June 15, 2022 (Table S1).

Statistical Analysis
We conducted a difference-in-differences analysis with staggered implementation to compare the
weekly incidence of COVID-19 in school districts that lifted mask requirements compared to school
districts where mask requirements had not yet been lifted.58–61 In this analysis, we estimated the weekly
and cumulative impact over 15 weeks of removing mask requirements on reported COVID-19 cases in
schools that removed masking requirements (i.e., average treatment effect among the treated).

Sensitivity analyses
We conducted various sensitivity analyses to ensure our results were robust to changes in model
specifications, data cleaning steps, and inclusion/exclusion criteria. Additional details can be found in
the Supplementary Appendix. Briefly, we varied data cleaning procedures (raw data vs. corrections
for 2-week reporting periods and corrections for non-reporting as zeroes), control groups (e.g.,
neighboring school districts only vs. entire NECTA division), weighting by schools’ population size,
smoothing/rolling average case rates, and adjustments for various covariates, including measures of
community burden of COVID-19 (Table S2). Our final analysis corrected for non-reporting as zeroes,
considered all school districts within the NECTA division as comparators, and was weighted by school
population size to capture the population impact of masking policies across our included school
districts. Our main analysis did not adjust for measures of community level COVID-19 burden, as a
growing body of evidence suggests schools as a driver of COVID-19 community burden,38,39,62 making it
a mediating factor along the causal pathway rather than a confounder of the relationship between
school masking policies and COVID-19 among students and staff.

Page 6 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Descriptive Analysis
Finally, to assess whether school masking policies were enacted with attention to health equity, we
compared timing of lifting/sustaining school masking policies across various school district
characteristics, including student and sociodemographics and physical characteristics of the learning
environment using data sources described above.

RESULTS

Out of the 72 included school districts in the Boston-Newton-Cambridge NECTA division, only Boston
and Chelsea Public Schools maintained masking throughout the study period (Figure 1A). Most school
districts (n=46, 64%) lifted masking requirements when the statewide mask requirement was rescinded
on February 28, 2022 (Figure 1B). The remaining 24 districts removed masking requirements in the
following reporting week (n=17, 24%) or two weeks (n=7, 10%) after the statewide mandate was lifted.

Prior to the lifting of masking requirements, as cases declined during the initial BA.1 Omicron wave,
reported COVID-19 case rates among students and staff were similar in Boston/Chelsea to school
districts within the study area that later lifted their masking requirements (Figure 2A). However,
following the lifting of masking requirements in late February through early March, reported COVID-19
rates diverged and were substantially higher in school districts that lifted their masking requirements
compared to those observed in Boston and Chelsea Public Schools where mask requirements were
sustained (Figure 2A). These trends held both overall (Figure 2A) and among students (Figure 2B)
and staff (Figure 2C) separately.

Next, we report individual weekly effects (Figure 3A) and cumulative effects over the study period
(Figure 3B) from our difference-in-differences analysis. We did not observe meaningful patterns in
differences between school districts in the period prior to the removal of the statewide school masking
requirement. Despite substantial variance between schools in the weeks corresponding to the initial
Omicron wave, the districts followed approximately the same trajectories during the surge with no clear
patterns that might violate the parallel trends assumption (Figure 3A.1). In contrast, once masking
requirements were lifted, we observed consistently higher case rates in those school districts compared
to districts which continued masking requirements. Lifting of masking requirements was associated with
significant increases in reported COVID-19 cases among students and staff in 12 out of 15 individual
weeks in the post-period (Figure 3A.1). These impacts increased with the amount of time since a

Page 7 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

school district had lifted their mask requirement. In the first week following a lifting of masking
requirements, we estimated an additional +1.44 (95% CI: +0.58, +2.29) COVID-19 cases per 1,000
students and staff, relative to school districts that had not yet lifted masking requirements. By the 9th
week after a school district had lifted mask requirements, +9.68 per 1,000 (+7.11, +12.25) excess
COVID-19 cases compared to school districts that had not yet lifted requirements (Figure 3A.1). We
observed similar trends between students and staff, with slightly stronger weekly effects observed
among staff (Figure 3A.2) compared to students (Figure 3A.3).

Importantly, the strength of the association between school masking requirements and COVID-19 case
rates varied with the background rate of COVID-19 in the surrounding community, such that the
strongest effects were observed in the weeks when background community case rates were at their
highest for a given school district (Figure S3, Figure S4). Determining whether this is due to high
community rates contributing to greater risk in unmasked schools or whether transmission in unmasked
schools is contributing to higher community rates is difficult and beyond the scope of this paper.
Nonetheless, these data clearly support a link between in-school transmission and community COVID-
19 rates.

In the 15 weeks following the lifting of the statewide masking requirement in schools, the cumulative
impact in school districts that lifted requirements increased over time for staff and students combined
(Figure 3B.1), and for students (Figure 3B.2) and staff (Figure 3B.3) separately. Overall, we estimate
that lifting of masking requirements was associated with an additional 44.9 (95% CI: 32.6, 57.1) cases
per 1,000 students and staff over the 15 weeks since the lifting of the statewide mandate (Table 1).
This excess rate corresponded to 11,901 (95% CI: 8,651, 15,151) total cases, representing 33.4%
(95% CI: 24.3%, 42.5%) of cases in school districts that lifted masking requirements and 29.4% (95%
CI: 21.4%, 37.5%) of cases in all school districts during that period.

This effect was even more pronounced among staff, where lifting requirements was associated with an
additional 81.7 (95% CI: 59.3, 104.1) COVID-19 cases per 1,000 staff over 15 weeks, representing
40.4% (95% CI: 29.4%, 51.5%) of all cases observed among staff in school districts that lifted masking
requirements and 35.1% (95% CI: 25.5%, 44.7%) of all cases observed among staff across all school
districts over this period. In sensitivity analyses, the results were robust to modeling specifications,
including data cleaning measures, population weighting, restricting comparison group school districts to
only neighboring school districts, and to adjusting for covariates including measures of community
burden of COVID-19 (Figure S5).

Page 8 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Finally, we found that school district characteristics varied with the length of time they sustained their
individual masking requirements following the lifting of the statewide mandate. We found that school
districts that sustained masking protections for longer periods were those with a higher percentage of
low-income students, students with disabilities, and English-learning students (Figure 4A) and a higher
percentage of Black and Latinx students (Figure 4B) and staff (Figure 4C). In addition, school districts
that sustained masking requirements for longer periods also had school buildings that were older, in
poorer physical condition (a category that included ventilation/HVAC), and higher numbers of students
per classroom (Figure 4D). In contrast, we found that school districts that lifted masking requirements
earlier tended to have a lower percentage of low income and English-learning students, students with
disabilities, Black and Latinx students and staff, fewer students per classroom, and newer buildings that
were generally in better condition. These differences between districts have important implications for
our results, as it clarifies that the increased COVID-19 rates in schools which removed masks are
unlikely to be related to a higher baseline risk of SARS-CoV-2 exposure for students and staff outside
of school. It also suggests that, despite newer buildings in better condition, ventilation in schools which
lifted mask requirements was generally not sufficient to substantially impact SARS-CoV-2 transmission

DISCUSSION

Given the dynamic nature of the pandemic, it is critical to examine in near real-time the need for and
impact of preventive measures, including masking in school settings. Schools are, and will continue to
be, an important yet politically contested space in the COVID-19 response, making analyses such as
this one particularly relevant to decision-makers. Our analysis adds to a growing body of literature
documenting the benefits of universal masking policies in public schools during a period of highly
transmissible SARS-CoV-2 variants and a rapidly changing COVID-19 policy environment.

Our paper documents significantly higher rates of reported COVID-19 in school districts that lifted
masking requirements compared to those that sustained masking requirements following the rescinding
of the Massachusetts statewide school masking order. Specifically, we estimate that rescinding mask
requirements in school districts in Eastern Massachusetts during March 2022 may have contributed an
additional 45 per 1,000 COVID-19 cases among students and staff in the 15 weeks following the end of
school-based masking requirements. In total, this represents more than 9,000 cases among students
and nearly 3,000 cases among staff. The Massachusetts Department of Elementary and Secondary
Education requires that children who test positive for COVID-19 isolate for a minimum of 5 days, or until
symptoms abate. In the best-case scenario, our results translate to a minimum of 17,505 days of school

Page 9 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

absence due to mandatory COVID-19 isolation in school children, and 6,547 days of teacher absence
over the 15 weeks since the lifting of the statewide mandate (see methods in Supplementary
Appendix, Table S3). Importantly, we observed the greatest impact of masking requirements in weeks
with highest background community rates of COVID-19, underscoring the importance of early
implementation and sustaining of school masking policies prior to and throughout surges. In addition,
given the evolving understanding of the impact of long-COVID on children,10–13 our results suggest that
masking requirements may be an important tool for school administrators and elected officials to
consider as they plan for the upcoming school year.

Understanding COVID-19 policy decisions requires attention to power and existing historical and
sociopolitical context.23,63–65 In the present study, we noted systematic differences in school masking
policy choices by school district characteristics. Specifically, school districts that sustained masking
requirements for longer periods tended to have school buildings in worse physical condition, more
crowded classrooms, and a higher proportion of staff and students made vulnerable by historic and
contemporary systems of oppression (e.g., racism, capitalism, xenophobia, and ableism). In addition,
the only two school districts to sustain school masking requirements through June – Boston and
neighboring Chelsea Public Schools – were also among the cities and towns in Massachusetts that
have been most impacted by the COVID-19 pandemic to date. These differences in length of school
masking policies may reflect an understanding among elected officials of how public policies are a key
mechanism by which structural racism operates to produce health inequities.23,30,33,63,65 Structural
racism and racial capitalism are fundamental causes of COVID-19 inequities.22,23 These forces operate
via diverse mechanisms such as household crowding, employment in essential industries, and access
to testing, vaccines and treatment, and differentially concentrate risk for both SARS-CoV-2 exposure
and severe COVID-19 outcomes among low income and Black, Latinx, and Indigenous communities.22–
24,37,66
Knowledge of these differential conditions may influence support/opposition to COVID-19
mitigation measures and policies in schools, including school masking.67,68 A growing body of work
suggests that knowledge of these inequities may result in increasing support for protective measures
among those directly impacted by structural racism and other systems of oppression while
simultaneously decreasing support among systematically advantaged groups whose relative position
largely insulates them from COVID-19 harms.67–69 In several studies and polls, Black and Latinx parents
were less likely to have confidence that schools could reopen safely without additional protections and
more likely to support school masking requirements.67,68,70,71 In contrast, when randomized to receive
information about pervasive racial/ethnic COVID-19 inequities, white individuals were less likely to
report concern about COVID-19, empathy for those vulnerable to COVID-19, and less likely to support
COVID-19 prevention policies.69 Given vastly unequal environmental conditions shaping COVID-19
Page 10 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

risk, universal school masking policies are an important measure for mitigating the impacts of structural
racism and may be especially important in settings where other preventive measures such as upgraded
ventilation/filtration may be more resource- and time-intensive to implement. School districts and
policymakers should consider these inequitable impacts when making plans about masking policies for
the upcoming school year.

Because universal school masking policies to prevent SARS-CoV-2 transmission has been a
contentious issue, we anticipate a number of critiques which we address here, several of which are
commonly levied at any study of masking in schools. One common critique of this type is that COVID-
19 is rare and mild in children. However, we observed weekly COVID-19 case rates exceeding 20 per
1,000 students at times during the study. These high rates likely represent substantial educational
disruptions and present increased risk of long-term complications (i.e., long-COVID). In addition, we
observed larger benefits of sustained masking among staff, who may be at increased risk of severe
COVID-19 outcomes. A second common critique of mandatory masking in schools is that alternative
approaches to reducing transmission, such as improved ventilation, exist. While this may be true in
theory, our findings make clear that the ventilation systems in Eastern Massachusetts school districts
were insufficient to prevent all COVID-19 cases in schools. Indeed, school districts with newer buildings
and better ventilation were more likely to remove masking requirements earlier, and thus our findings of
increased COVID-19 incidence in the absence of masks demonstrate that ventilation likely remained
insufficient in most schools in our sample, reinforcing the need for layered mitigation measures.

A key strength of this study is our use of staggered dates of removal of mask requirements, and
difference-in-differences methodology. These approaches enabled us to estimate the impact of
masking requirements despite differences between school districts. While it is true that there are
several factors which differ between school districts, and which are related to SARS-CoV-2 exposure
risk, difference-in-differences methodology is immune to those sources of confounding whenever they
do not experience changes over time coinciding with the policy change of interest (e.g., differences in
sociodemographics or vaccination rates). Furthermore, when we investigated the distribution of known
COVID-19 risk factors between school districts, we found that those districts which removed masks
were those which a priori would have been expected to have lower COVID-19 rates. This suggests that
any residual confounding of our results by COVID-19 risk would have led to our analysis
underestimating the harms of removing universal masking policies.

A key question in any study looking at schools is how background community COVID-19 rates interact
with school-specific rates. There are two potential reasons why community levels may be relevant, and
Page 11 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

the correct analytic decision on whether to include community levels as confounder or not, is not
straightforward. In sensitivity analyses, we found that the benefits of universal masking requirements
persisted even after controlling for several measures of community COVID-19 incidence. We, however,
did not prioritize these results in our main analyses, as we argue it is more appropriate to consider
community rates of COVID-19 as part of the causal effect of school masking policies rather than a
source of bias. School and community COVID-19 levels are so closely linked that it is difficult to rule out
at least some of the variation in community COVID-19 rates being a direct consequence of changing
school case rates.63 That is, students and staff can infect their family and community members and vice
versa. In addition, while some community-level policies changed during Spring 2022, these changes did
not perfectly coincide with the school mask mandates being rescinded, nor did the community mask
policies always align with school mask policies. Overall, we anticipate that although community case
rates change over time and are connected to school case rates, the ways in which community cases
and school cases are related should not depend on school masking policies.

A second reason to consider community COVID-19 cases is that there may be spillover between school
districts or communities as individuals move between areas. This is not an issue of confounding, but
rather a potential threat to the validity of estimating causal effects of COVID-19 prevention policies.
Spillover between communities and school districts may reduce the difference in COVID-19 rates and
has the potential to alter the case growth trajectory in masked schools. This latter issue is a potential
threat to the parallel trends assumption required for difference-in-differences analyses. To address this,
we used the staggered policy adoption model of difference-in-differences.59–62 This approach uses all
pre-policy change data as unexposed data, and aligns the comparisons for school districts which
unmask to contemporaneous control districts, allowing us to estimate the impact of removing masks,
uncoupled from calendar time. As a result, we anticipate that any remaining impact of spillover on our
findings would be to decrease the estimated impact of removing masks, making our results an
underestimate of the harms caused by removing mask mandates earlier.

Overall, our findings should be interpreted as the impact of universal masking policies, not masks per
se, given that it is unlikely that all children and families removed masks when the requirement was
removed, as masks were still encouraged in most school settings and utilized by many. Despite this,
the impact of lifting required masking policies was substantial.

Page 12 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

CONCLUSION

Undoubtedly, the Omicron wave will not be the final COVID-19 surge, and ongoing efforts to minimize
the impacts of COVID-19 in school settings, including evidence-based public policy, are urgently
needed. Our results underscore the importance of centering health equity in these policy choices and
support universal school masking policies as an important piece of a layered mitigation strategy to
reduce COVID-19 risk among those made vulnerable by structural racism and other systems of
oppression. If trends from prior years persist, surges may be especially likely to occur in late December
and January. Many students and staff will likely have spent time traveling, gathering indoors, or in other
high-risk transmission settings during winter break, and epidemiologic best practices suggest a 14-day
quarantine following these types of activities to prevent onward transmission. We recommend that
school districts develop mitigation plans proactively in anticipation of a winter COVID-19 wave during
the 2022-23 school year. In particular, requiring masks in schools in December and January, with a
clear a priori decision threshold for removing masks in March or later as the winter wave abates, could
be an effective strategy to minimize the impact of COVID-19 in school settings.

Page 13 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

REFERENCES
1. Clarke KEN, Jones JM, Deng Y, et al. Seroprevalence of Infection-Induced SARS-CoV-2 Antibodies -
United States, September 2021-February 2022. MMWR Morb Mortal Wkly Rep 2022;71(17):606–8.

2. Iuliano AD, Brunkard JM, Boehmer TK, et al. Trends in Disease Severity and Health Care Utilization
During the Early Omicron Variant Period Compared with Previous SARS-CoV-2 High Transmission
Periods - United States, December 2020-January 2022. MMWR Morb Mortal Wkly Rep
2022;71(4):146–52.

3. Payne AB, Gilani Z, Godfred-Cato S, et al. Incidence of Multisystem Inflammatory Syndrome in

Children Among US Persons Infected With SARS-CoV-2. JAMA Netw Open 2021;4(6):e2116420.

4. Abrams JY, Oster ME, Godfred-Cato SE, et al. Factors linked to severe outcomes in multisystem
inflammatory syndrome in children (MIS-C) in the USA: a retrospective surveillance study. Lancet Child
Adolesc Health 2021;5(5):323–31.

5. McCormick DW, Richardson LC, Young PR, et al. Deaths in Children and Adolescents Associated With
COVID-19 and MIS-C in the United States. Pediatrics [Internet] 2021;148(5). Available from:
http://dx.doi.org/10.1542/peds.2021-052273

6. Godfred-Cato S, Bryant B, Leung J, et al. COVID-19-Associated Multisystem Inflammatory Syndrome in

Children - United States, March-July 2020. MMWR Morb Mortal Wkly Rep 2020;69(32):1074–80.

7. Feldstein LR, Tenforde MW, Friedman KG, et al. Characteristics and Outcomes of US Children and
Adolescents With Multisystem Inflammatory Syndrome in Children (MIS-C) Compared With Severe
Acute COVID-19. JAMA 2021;325(11):1074–87.

8. Wanga V, Gerdes ME, Shi DS, et al. Characteristics and Clinical Outcomes of Children and
Adolescents Aged <18 Years Hospitalized with COVID-19 - Six Hospitals, United States, July-August
2021. MMWR Morb Mortal Wkly Rep 2021;70(5152):1766–72.

9. Miller, Yousaf, Bornstein. Multisystem Inflammatory Syndrome in Children (MIS-C) During SARS-CoV-2
Delta and Omicron Variant Circulation—United States, July 2021–January 2022. Clin Infect Dis
[Internet] 2022;Available from: https://academic.oup.com/cid/advance-article-
abstract/doi/10.1093/cid/ciac471/6605071

10. CDC. Long COVID or Post-COVID Conditions [Internet]. Centers for Disease Control and Prevention.
2022 [cited 2022 Jul 5];Available from: https://www.cdc.gov/coronavirus/2019-ncov/long-term-
effects/index.html

11. Stephenson T, Pinto Pereira SM, Shafran R, et al. Physical and mental health 3 months after SARS-
CoV-2 infection (long COVID) among adolescents in England (CLoCk): a national matched cohort
study. Lancet Child Adolesc Health 2022;6(4):230–9.

12. Lopez-Leon S, Wegman-Ostrosky T, Ayuzo del Valle NC, et al. Long-COVID in children and
adolescents: A systematic review and meta-analyses [Internet]. bioRxiv. 2022;Available from:
https://www.medrxiv.org/content/10.1101/2022.03.10.22272237.abstract

13. Kikkenborg Berg S, Palm P, Nygaard U, et al. Long COVID symptoms in SARS-CoV-2-positive children
aged 0-14 years and matched controls in Denmark (LongCOVIDKidsDK): a national, cross-sectional
study. Lancet Child Adolesc Health [Internet] 2022;Available from: http://dx.doi.org/10.1016/S2352-
4642(22)00154-7

14. Magnusson K, Skyrud KD, Suren P, et al. Healthcare use in 700 000 children and adolescents for six
months after covid-19: before and after register based cohort study. BMJ 2022;376:e066809.
Page 14 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

15. Asadi-Pooya AA, Nemati H, Shahisavandi M, et al. Long COVID in children and adolescents. World J
Pediatr 2021;17(5):495–9.

16. Zimmermann P, Pittet LF, Curtis N. How Common is Long COVID in Children and Adolescents? Pediatr
Infect Dis J 2021;40(12):e482–7.

17. Izquierdo-Pujol J, Moron-Lopez S, Dalmau J, et al. Post COVID-19 Condition in Children and
Adolescents: An Emerging Problem. Front Pediatr 2022;10:894204.

18. Kompaniyets L, Bull-Otterson L, Boehmer TK, et al. Post-COVID-19 Symptoms and Conditions Among
Children and Adolescents - United States, March 1, 2020-January 31, 2022. MMWR Morb Mortal Wkly
Rep 2022;71(31):993–9.

19. Hillis SD, Blenkinsop A, Villaveces A, et al. COVID-19-Associated Orphanhood and Caregiver Death in
the United States. Pediatrics [Internet] 2021;Available from: http://dx.doi.org/10.1542/peds.2021-
053760

20. Parks SE, Zviedrite N, Budzyn SE, et al. COVID-19-Related School Closures and Learning Modality
Changes - United States, August 1-September 17, 2021. MMWR Morb Mortal Wkly Rep
2021;70(39):1374–6.

21. Nunez-Smith M, COVID-19 Health Equity Task Force. Presidential COVID-19 Health Equity Task Force
- Final Report and Recommendations [Internet]. U.S. Department of Health and Human Services; 2021.
Available from:
https://www.minorityhealth.hhs.gov/assets/pdf/HETF_Report_508_102821_9am_508Team%20WIP11.
pdf

22. Laster Pirtle WN. Racial Capitalism: A Fundamental Cause of Novel Coronavirus (COVID-19)
Pandemic Inequities in the United States. Health Educ Behav 2020;47(4):504–8.

23. Bailey ZD, Moon JR. Racism and the Political Economy of COVID-19: Will We Continue to Resurrect
the Past? J Health Polit Policy Law 2020;45(6):937–50.

24. Emerson MA, Montoya T. Confronting Legacies of Structural Racism and Settler Colonialism to
Understand COVID-19 Impacts on the Navajo Nation. Am J Public Health 2021;111(8):1465–9.

25. Cowger TL, Davis BA, Etkins OS, et al. Comparison of Weighted and Unweighted Population Data to
Assess Inequities in Coronavirus Disease 2019 Deaths by Race/Ethnicity Reported by the US Centers
for Disease Control and Prevention. JAMA Netw Open 2020;3(7):e2016933.

26. Shi DS, Whitaker M, Marks KJ, et al. Hospitalizations of Children Aged 5-11 Years with Laboratory-
Confirmed COVID-19 - COVID-NET, 14 States, March 2020-February 2022. MMWR Morb Mortal Wkly
Rep 2022;71(16):574–81.

27. Bailey LC, Razzaghi H, Burrows EK, et al. Assessment of 135 794 Pediatric Patients Tested for Severe
Acute Respiratory Syndrome Coronavirus 2 Across the United States. JAMA Pediatr 2021;175(2):176–
84.

28. Marks KJ, Whitaker M, Anglin O, et al. Hospitalizations of Children and Adolescents with Laboratory-
Confirmed COVID-19 - COVID-NET, 14 States, July 2021-January 2022. MMWR Morb Mortal Wkly
Rep 2022;71(7):271–8.

29. Xiao Y, Yip PS-F, Pathak J, Mann JJ. Association of Social Determinants of Health and Vaccinations
With Child Mental Health During the COVID-19 Pandemic in the US. JAMA Psychiatry 2022;79(6):610–
21.

Page 15 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

30. Neal DE. HEALTHY SCHOOLS: A MAJOR FRONT IN THE FIGHT FOR ENVIRONMENTAL JUSTICE.
Environ Law 2008;38(2):473–93.

31. Wilson S, Hutson M, Mujahid M. How Planning and Zoning Contribute to Inequitable Development,
Neighborhood Health, and Environmental Injustice. Environ Justice 2008;1(4):211–6.

32. Bullard RD. Dismantling Environmental Racism in the USA. Local Environ 1999;4(1):5–19.

33. Bailey ZD, Krieger N, Agénor M, Graves J, Linos N, Bassett MT. Structural racism and health inequities
in the USA: evidence and interventions. Lancet 2017;389(10077):1453–63.

34. Bullard RD, Wright BH. Environmental justice for all: community perspectives on health and research
needs. Toxicol Ind Health 1993;9(5):821–41.

35. Bullard RD. The Threat of Environmental Racism. Nat Resour Environ 1993;7(3):23–56.

36. Pastor M Jr, Sadd JL, Morello-Frosch R. Who’s minding the kids? Pollucion, public schools, and
environmental justice in Los Angeles. Soc Sci Q 2002;83(1):263–80.

37. Kitzmiller EM, Drake Rodriguez A. Addressing Our Nation’s Toxic School Infrastructure in the Wake of
COVID-19. Educ Res 2022;51(1):88–92.

38. Wiens KE, Smith CP, Badillo-Goicoechea E, et al. In-person schooling and associated COVID-19 risk in
the United States over spring semester 2021. Sci Adv 2022;8(16):eabm9128.

39. Lessler J, Grabowski MK, Grantz KH, et al. Household COVID-19 risk and in-person schooling. Science
2021;372(6546):1092–7.

40. Gettings J, Czarnik M, Morris E, et al. Mask Use and Ventilation Improvements to Reduce COVID-19
Incidence in Elementary Schools - Georgia, November 16-December 11, 2020. MMWR Morb Mortal
Wkly Rep 2021;70(21):779–84.

41. Falk A, Benda A, Falk P, Steffen S, Wallace Z, Høeg TB. COVID-19 Cases and Transmission in 17 K-
12 Schools - Wood County, Wisconsin, August 31-November 29, 2020. MMWR Morb Mortal Wkly Rep
2021;70(4):136–40.

42. Budzyn SE, Panaggio MJ, Parks SE, et al. Pediatric COVID-19 Cases in Counties With and Without
School Mask Requirements - United States, July 1-September 4, 2021. MMWR Morb Mortal Wkly Rep
2021;70(39):1377–8.

43. Jehn M, McCullough JM, Dale AP, et al. Association Between K-12 School Mask Policies and School-
Associated COVID-19 Outbreaks - Maricopa and Pima Counties, Arizona, July-August 2021. MMWR
Morb Mortal Wkly Rep 2021;70(39):1372–3.

44. Donovan CV, Rose C, Lewis KN, et al. SARS-CoV-2 Incidence in K-12 School Districts with Mask-
Required Versus Mask-Optional Policies - Arkansas, August-October 2021. MMWR Morb Mortal Wkly
Rep 2022;71(10):384–9.

45. Andrejko KL, Pry JM, Myers JF, et al. Effectiveness of Face Mask or Respirator Use in Indoor Public
Settings for Prevention of SARS-CoV-2 Infection - California, February-December 2021. MMWR Morb
Mortal Wkly Rep 2022;71(6):212–6.

46. Murray TS, Malik AA, Shafiq M, et al. Association of Child Masking With COVID-19-Related Closures in
US Childcare Programs. JAMA Netw Open 2022;5(1):e2141227.

Page 16 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

47. Boutzoukas AE, Zimmerman KO, Inkelas M, et al. School Masking Policies and Secondary SARS-CoV-
2 Transmission. Pediatrics [Internet] 2022;Available from: http://dx.doi.org/10.1542/peds.2022-056687

48. Dawson P, Worrell MC, Malone S, et al. Pilot Investigation of SARS-CoV-2 Secondary Transmission in
Kindergarten Through Grade 12 Schools Implementing Mitigation Strategies - St. Louis County and City
of Springfield, Missouri, December 2020. MMWR Morb Mortal Wkly Rep 2021;70(12):449–55.

49. CDC. Guidance for COVID-19 Prevention in K-12 Schools [Internet]. Centers for Disease Control and
Prevention. 2021 [cited 2022 Jul 5];Available from:
https://web.archive.org/web/20211001030138/https://www.cdc.gov/coronavirus/2019-
ncov/community/schools-childcare/k-12-guidance.html

50. CDC. COVID-19 Community Levels [Internet]. Centers for Disease Control and Prevention. 2022 [cited
2022 Jul 27];Available from: https://www.cdc.gov/coronavirus/2019-ncov/science/community-levels.html

51. Camera L. New CDC Mask Guidance Marks Major Shift for Schools [Internet]. U.S. News & World
Report. 2022 [cited 2022 Jul 5];Available from: https://www.usnews.com/news/national-
news/articles/2022-02-25/new-cdc-mask-guidance-marks-major-shift-for-schools

52. Decker S. Which States Ban Mask Mandates in Schools, and Which Require Masks? [Internet].
Education Week. 2021 [cited 2022 Jul 5];Available from: https://www.edweek.org/policy-politics/which-
states-ban-mask-mandates-in-schools-and-which-require-masks/2021/08

53. Massachusetts Department of Elementary and Secondary Education. Positive COVID-19 Cases in
Schools [Internet]. [cited 2022 Jun 22];Available from: https://www.doe.mass.edu/covid19/positive-
cases/default.html#weekly-report

54. School and District Profiles - Statewide Reports [Internet]. Massachusetts Department of Elementary
and Secondary Education (DESE). Available from: https://profiles.doe.mass.edu/state_report/

55. DESE/DPH Protocols for Responding to COVID-19 Scenarios - SY 2021-22 [Internet]. 2022 [cited 2022
Jun 22];Available from: https://www.doe.mass.edu/covid19/on-desktop/protocols/protocols.pdf

56. Massachusetts Department of Elementary and Secondary Education (DESE), Massachusetts

Department of Public Health (DPH). New COVID-19 Testing Program Option for K-12 Schools
[Internet]. 2022 [cited 2022 Jun 22];Available from: https://www.doe.mass.edu/covid19/on-
desktop/2022-0118new-testing-program.pdf

57. Massachusetts School Building Authority. 2016 SCHOOL SURVEY REPORT [Internet]. 2016. Available
from: https://www.massschoolbuildings.org/sites/default/files/edit-
contentfiles/Programs/School_Survey/2016/MSBA%202016%20Survey%20Report_102417-FINAL.pdf

58. Athey S, Imbens GW. Design-based analysis in Difference-In-Differences settings with staggered
adoption. J Econom 2022;226(1):62–79.

59. Goodman-Bacon A. Difference-in-differences with variation in treatment timing. J Econom

2021;225(2):254–77.

60. Callaway B, Sant’Anna PHC. Difference-in-Differences with multiple time periods. J Econom
2021;225(2):200–30.

61. Roth J, Sant’Anna PHC, Bilinski A, Poe J. What’s Trending in Difference-in-Differences? A Synthesis of
the Recent Econometrics Literature [Internet]. arXiv [econ.EM]. 2022;Available from:
http://arxiv.org/abs/2201.01194

Page 17 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

62. Johnson KE, Lachmann M, Stoddard M, et al. Detecting in-school transmission of SARS-CoV-2 from
case ratios and documented clusters. medRxiv [Internet] 2021;Available from:
http://dx.doi.org/10.1101/2021.04.26.21256136

63. Michener J. Race, power, and policy: understanding state anti-eviction policies during COVID-19. Policy
Soc 2022;41(2):231–46.

64. Woolhandler S, Himmelstein DU, Ahmed S, et al. Public policy and health in the Trump era. Lancet
2021;397(10275):705–53.

65. Bailey ZD, Feldman JM, Bassett MT. How Structural Racism Works - Racist Policies as a Root Cause
of U.S. Racial Health Inequities. N Engl J Med 2021;384(8):768–73.

66. Krieger N, Waterman PD, Chen JT. COVID-19 and Overall Mortality Inequities in the Surge in Death
Rates by Zip Code Characteristics: Massachusetts, January 1 to May 19, 2020. Am J Public Health
2020;110(12):1850–2.

67. Cotto R Jr, Woulfin S. Choice With(out) Equity? Family Decisions of Child Return to Urban Schools in
Pandemic. JFDE 2021;4(1):42–63.

68. Gilbert LK, Strine TW, Szucs LE, et al. Racial and Ethnic Differences in Parental Attitudes and
Concerns About School Reopening During the COVID-19 Pandemic - United States, July 2020. MMWR
Morb Mortal Wkly Rep 2020;69(49):1848–52.

69. Skinner-Dorkenoo AL, Sarmal A, Rogbeer KG, André CJ, Patel B, Cha L. Highlighting COVID-19 racial
disparities can reduce support for safety precautions among White U.S. residents. Soc Sci Med
2022;301:114951.

70. Annie Ma, Associated Press, Hannah Fingerhut, Associated Press. Racial split on COVID-19 endures
as restrictions ease in U.S [Internet]. PBS NewsHour. 2022 [cited 2022 May 17];Available from:
https://www.pbs.org/newshour/health/racial-split-on-covid-19-endures-as-restrictions-ease-in-us

71. Parr R. New K-12 Parent Poll: Student mental health, academics pose challenge to Massachusetts
schools COVID recovery plans [Internet]. MassINC Polling Group. 2022 [cited 2022 Aug 1];Available
from: https://www.massincpolling.com/the-topline/new-k-12-parent-poll-student-mental-health-
academics-pose-challenge-to-massachusetts-schools-covid-recovery-plans

Page 18 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

TABLES & FIGURES

Figure 1. (A) Map of school districts in Boston-Cambridge-Newton NECTA division by reporting week
in which mask requirements were removed and (B) number of school districts with mask requirements
remaining in place over time1

1
Black dots denote the n=7 school districts excluded from the analysis due to unreliable data reporting, see
Figure S1
Page 19 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Figure 2. Weekly reported rate of COVID-19 cases (A) overall, (B) among students, and (C) among
staff in Boston/Chelsea Public Schools, by week masking requirements were lifted for school districts
within the Boston-Cambridge-Newton NECTA division.2

2
Dates on the x-axis were restricted to the period immediately before and after universal school masking
requirements were lifted statewide and in most school districts. Difference-in-differences analysis includes all
weeks in the 2021-2022 school year.
Page 20 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Figure 3. Weekly (A) and cumulative (B) differences in rate of COVID-19 overall (top), among students
(middle) and among staff (bottom) in school districts that lifted masking requirements compared to
school districts that had not yet lifted their masking requirements in that week.3

3
Results are modeled estimates from our difference-in-differences analysis. The grey band in the background of
the plot depicts the initial BA.1 Omicron wave in December 2021-January 2022.
Page 21 of 23
Table 1. Rates of COVID-19 overall and by school district masking policies and estimated impact of lifting masking requirements in the 15
calendar weeks following lifting of statewide school masking requirements, overall and for staff and students separately
Mask Mask
All
Requirements Requirements Average effect of lifting mask requirements
Districts
Sustained Lifted
Rate Rate Rate Attributable
n Avg. n n Rate Diff. AF% PAF%
Group per per per cases
Cases Pop Cases Cases (95% CI)b (95% CI)d (95% CI)e
1,000a 1,000 1,000 (95% CI)c
Overall 44.9 11,901 33.4% 29.4%
40,416 337,226 119.8 4,766 66.1 35,651 134.4
(Staff & Students) (32.6, 57.1) (8,651, 15,151) (24.3%, 42.5%) (21.4%, 37.5%)
39.9 9,168 32.1% 28.5%
Students 32,198 291,149 110.6 3,674 60.0 28,524 124.1
(24.3, 55.4) (5,594, 12,743) (19.6%, 44.7%) (17.4%, 39.6%)
81.7 2,882 40.4% 35.1%
Staff 8,218 46,077 178.4 1,091 101.0 7,127 202.1
(59.3, 104.1) (2,092, 3,673) (29.4%, 51.5%) (25.5%, 44.7%)
a
All rates represent cumulative reported COVID-19 cases per 1,000 population over the 15 calendar weeks since DESE lifted the statewide school masking requirement
on 28-Feb-2022
b
From difference-in-differences models, the difference in COVID-19 rates per 1,000 attributable to lifting masking policies in districts that lifted masking requirements,
compared to districts that maintained masking requirements over the 15 weeks since the statewide school masking requirement was lifted. The cumulative average
treatment effect among the treated (ATT).
c
Total number of COVID-19 cases attributable to lifting masking requirements in the school districts that lifted masking requirements over the 15 calendar weeks post-lifting
of statewide school masking requirements. Calculated by multiplying cumulative ATT by the average number of students in school districts that removed masking
requirements.
d
Attributable fraction (expressed as a percent): the percentage of COVID-19 cases among students in schools that did not lift mask mandates attributable to lifting masking
requirements over the 15 calendar weeks post-lifting of statewide school masking requirements. Calculated by taking the cumulative ATT divided by the observed
COVID-19 case rate in school districts that lifted masking requirements
e
Population attributable fraction (PAF%), expressed as a percentage: the percentage of COVID-19 cases in all school districts - both those that lifted and those that
sustained masking requirements - attributable to lifting masking requirements over the 15 calendar weeks post-lifting of statewide school masking requirements.
Calculated by taking number of attributable cases divided by the total number of cases observed across all districts

Page 22 of 23
medRxiv preprint doi: https://doi.org/10.1101/2022.08.09.22278385; this version posted August 9, 2022. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Figure 4. Distribution4 of selected characteristics of school districts by week masking requirements

were lifted, including (A) percentage of students in selected populations as defined by the
Massachusetts Department of Elementary and Secondary Education (DESE), race/ethnicity of (B)
students and (C) staff, and (D) physical building conditions and learning environment characteristics
from the 2016-2017 Massachusetts School Building Authority (MSBA) school survey

4
Variables are scaled to enable them to be depicted on the same scale – zero represents the mean value with
units in standard deviations
Page 23 of 23