The Politics of School Reopening During COVID-19: A Multiple Case Study of Five Urban Districts in the 2020–21 School Year

Abstract

Purpose: Nearly all schools in the United States closed in spring 2020, at the onset of the COVID-19 pandemic. We analyze traditional public and charter school reopenings for the 2020–21 school year in five urban districts. We provide a rich and theoretically grounded description of how and why educational leaders made reopening decisions in each of our case districts. Research Methods: We used data from a multiple-case study from March 2020 to July 2021. The research team conducted 56 interviews with school, district, and system-level leaders; triangulated with publicly available data; and also drew on interview data from a subsample of parents and guardians in each of our sites. We analyzed these data through qualitative coding and memo writing and conducted detailed single- and cross-case analyses. Findings: School system leaders in our case sites generally consulted public health authorities, accounted for state-level health and educational guidance, and engaged with and were responsive to the interests of different stakeholders. Districts’ adherence to and strategic uses of public health guidance, as well as a combination of union-district relations and labor market dynamics, influenced reopening. Parents, city, and state lawmakers, and local institutional conditions also played a role, helping to explain differences across cases. Implications: In contrast to the “politics or science” framing that has dominated research and public discourse on school reopening, we show that local pandemic conditions and local political dynamics both mattered and in fact were interrelated. Our findings have some implications for how educational leaders might navigate future crises.

Introduction

Nearly all schools in the United States closed and ceased in-person learning in spring 2020, at the onset of the COVID-19 pandemic (Grossmann et al., 2021). For the 2020–21 school year, educational leaders needed to decide when and how to “reopen” their schools, or reintroduce in-person learning options. These decisions were politically and ethically difficult, as leaders weighed the social, emotional, and academic consequences for students along with the health threat to students, parents, educators, and the community at large (Freidus & Turner, 2022). ¹ So, how did educational leaders decide when and how to reopen their schools, and what factors influenced their decisions? An initial wave of research has identified how COVID-19 rates and a number of political and demographic factors are associated with reopening status. Yet the existing studies offer limited descriptions of the process of reopening and limited explanations for how various factors (e.g., COVID-19 rates, partisanship, union strength, demographics) led to reopening decisions (Singer, 2022). This makes interpretations of existing findings on school reopening during the pandemic less clear and limits their usefulness for informing educational decision-making in future crises.

In this study, we examine how traditional public and charter schools reopened in five urban districts. Considering the reopening process through the lens of micropolitics, we describe the actors involved; the political, institutional, cultural, and public health contexts in which they operated; and the way they interacted in the decision-making arena. Specifically, we answer the following research questions:

1. How did public school districts and charter schools reopen in Denver, Detroit, New Orleans, Portland, and Washington, D.C., in 2020–21?

2. What factors shaped the school reopening decisions in each of these districts?

   1. Who were the key actors and what strategies did they use to influence decision-making?
   2. How were these actors and their interactions influenced by their political, institutional, cultural, and public health contexts?

First, we offer a rich description of reopening decisions in each of our case districts. School system leaders in our case sites generally consulted public health authorities, accounted for state-level health and educational guidance, and engaged with and were responsive to the interests of different stakeholders (e.g., teachers and teachers’ unions, parents). They constructed plans and timelines for reopening before the start of the school year. Although plans differed between our case sites, as did the local conditions at the start of the school year and over time, our case districts generally planned for delayed reopenings. They largely held to these plans throughout the year, even as COVID-19 case-rates later rose to their highest levels yet in some of our case sites. The way that schools reopened also differed between our case sites, though an important commonality was that schools continued to offer students a remote option. Our data also suggest that leaders traditional public school (TPS) and charter schools within the same districts had similar reopening processes and timelines.

Second, we offer theoretically grounded explanations for how factors identified in prior studies influenced reopening decisions. Our micropolitical analysis helps illustrate why partisanship mattered: in these five heavily Democratic districts, school and system leaders adhered to and strategically drew upon public health guidance to legitimate their decisions and avoid risk, and they perceived a great deal of caution and health concerns among their families. Our findings also clarify the role of teachers and unions: in some cases, unions used specific tactics to influence decision-making, but both unionized TPS and largely nonunionized charters were generally attentive to teacher well-being and influenced by concerns about teacher retention. We also highlight the role of state lawmakers and the local institutional context. The extent to which these factors varied between our case sites helps explain why the nature and timing of their reopening differed. These findings provide a clearer picture of how and why districts reopened in 2020–21—the first full school year of the pandemic.

Literature Review

A number of studies have examined school reopening during the pandemic, addressing the question, “What factors are associated with school reopening in the 2020–21 school year?” These studies have offered some consistent findings as well as some mixed results, highlighting the need for more research. Ultimately, while these studies have identified factors associated with reopening, they have not provided clear evidence on how these factors shaped the nature and timing of reopening or how their impact was interrelated (Singer, 2022).

Two political factors—partisanship and teachers’ union strength—are strong predictors of reopening status. Districts with more Democratic support were less likely to offer in-person instruction (Grossmann et al., 2021; Harris & Oliver, 2021; Hartney & Finger, 2021; Houston & Steinberg, 2022; Marianno et al., 2022). Kretchmar and Brewer's (2022) qualitative case study of reopening in Wisconsin and Georgia districts show how the partisan affiliation of governing elites and constituents translated into different reopening decisions. With a politicized response to public health guidance, there were partisan differences in how Democratic and Republican districts adopted and adhered to public health guidance. Districts with stronger teachers' unions were also less likely to offer in-person instruction (DeAngelis & Makridis, 2021; Grossmann et al., 2021; Harris & Oliver, 2021; Hartney & Finger, 2021; Houston & Steinberg, 2022). Marianno et al. (2022) find that unions in 250 large school districts had influence through their “second face of power” (i.e., existing influence through bargaining and negotiation power) but not their “first face of power” (i.e., efforts to influence decisions through new organizing).

Community demographics also appear relevant. Many studies use district demographics (e.g., race or ethnicity) as a control and do not report them as a factor in reopening. Those that do typically find greater shares of Black, Hispanic, and low-income families predict a lower likelihood of reopening (Diemer & Park, 2022; Grossmann et al., 2021; Haderlein et al., 2021b; Harris & Oliver, 2021; Hartney & Finger, 2021; Marianno et al., 2022), though at least one study finds that they are not statistically significant (Houston & Steinberg, 2022). A substantial body of research has also assessed whether and why there are racial and socioeconomic gaps in preferences for and participation in in-person instruction (e.g., Calarco et al., 2021; Camp & Zamarro, 2021; Collins, 2021a; Cotto Jr. & Woulfin, 2021; Darling-Aduana et al., 2022; Grossmann et al., 2021; Haderlein et al., 2021a, 2021b; Harris & Oliver, 2021; Kogan, 2021; Parolin & Lee, 2021).

Findings on the impact of COVID-19 itself are mixed, but this may be a function of the way that researchers operationalize this factor. A large share of studies found that COVID-19 rates were not a statistically significant predictor of reopening decisions (Singer, 2022). Houston and Steinberg (2022), however, show that when examining variation between districts within states, “counties with higher Covid case rates had lower in-person instruction rates” (p. 19). In addition, Christian et al. (2022) found that in Ohio, marginal increases in a school's COVID-19 cases reduced the probability that school would be in-person the next week. Thus, earlier studies’ reliance on county-level data and national (rather than within-state) comparisons may explain why they found COVID-19 rates to be insignificant predictors. Finally, Harris and Oliver (2021) also emphasize that prior studies may muddle the ways in which prior and current COVID-19 conditions influenced and interacted with these other factors.

Evidence on the impact of market competition and governance structure is also mixed. Some studies find an impact of private school competition on reopening, while others did not; and charter school competition does not appear to have impacted reopening (Harris & Oliver, 2021; Hartney & Finger, 2021; Marianno et al., 2022). Charter schools appear to follow similar patterns and have similar contextual and constituent influences as TPS (Cohodes & Pitts, 2021; Camp & Zamarro, 2021; Grossmann et al., 2021), though Grossmann et al. (2021) show Michigan charters were more likely to plan for in-person instruction in majority-Black districts.

There are three limitations in this research that our study seeks to address. First, the research on school reopening has been largely based on correlational analyses of large-scale quantitative data sets (Singer, 2022). The studies offer a broad perspective on school responses to the COVID-19 pandemic, which is beneficial since research on school leadership during crises is often single-site case studies with isolated or context-dependent findings (Grissom & Condon, 2021). Although the quantitative studies have identified some strong correlates with reopening during the COVID-19 pandemic, however, they have not clearly identified how or why these factors led to the reopening decisions they are studying. Our study details the contexts and processes hinted at by the association between these political and demographic factors and school reopening (Maxwell, 2012).

Second, the existing research defines reopening based on categories (e.g., in-person, hybrid, remote) that gloss over differences in implementation, and often only analyzes these reopening categories for one or two time periods. Our analysis captures reopening as a dynamic process and provides some additional detail about the nature of the reopening process and differences in the implementation of in-person instruction. Finally, the existing research largely neglects charter schools, in part due to the lack of charters in readily available data sets, even though charters serve a substantial number of low-income and racially minoritized students, who have been most negatively impacted by the pandemic (Hardy & Logan, 2020; Kuhfeld et al., 2022; Magesh et al., 2021; National Center for Education Statistics, 2021). With data from some charter schools in each of our case sites, we address whether and why charter leaders responded similarly or differently than TPS leaders in the same districts.

Conceptual Framework of Micropolitics

We ground our analysis in a micropolitical framework that focuses on the behaviors and values of actors, the resources they have to influence policy decisions, and their interactions with others and their local context (Iannaccone, 1991). According to Blase (1991) “micropolitics” refers to “formal and informal power by individuals and groups to achieve their goals in organizations”:

In large part, political actions result from perceived differences between individuals and groups, coupled with the motivation to use power to influence and/or protect. Although such actions are consciously motivated, any action, consciously or unconsciously motivated, may have political “significance” in a given situation. Both cooperative and conflictive actions and processes are part of the realm of micropolitics. (p. 11)

This framework acknowledges that policy adoption and implementation depend greatly on the interaction of actors (and coalitions) who draw on their positions of authority (formal power) and on their social, cultural, and intellectual capital (informal power) to support and resist policies in ways that advance their individual and collective interests. These political maneuvers are also mediated by the social environment to determine whether a policy gets on the agenda and is adopted and whether it is implemented in ways that achieve its goals (Malen, 2006; Marsh, 2012). This framework is particularly suitable for our multicase analysis because it (1) considers the organization as the unit of analysis (in our case, school systems and individual schools), and (2) zooms in on the political strategies and decisions of different actors and how these actions interact with the local context (in our case, communities with varying sociopolitical environments)—allowing for nuanced comparisons across our cases.

Micropolitics suggests that the interaction of the policy, the actors, and the context explains whether that policy gets adopted and how it gets implemented. According to this framework:

• When the underlying assumptions, purpose, and values of a policy option aligns with actors’ interests and values, they will be more motivated to marshal resources in support of its adoption and implementation. Conversely, misalignment creates opportunities for political resistance and incentives to amend it (Marsh, 2012). In our study, the policy options to reopen schools in person and in what form tapped into deeply held values around safety, personal liberty, and equity, creating ample opportunity for disagreement and conflict.

• Actors come to the process “seek[ing] to promote and protect their vested material and ideological interests” and to advance their preferences or conceptions of the “public good” (Malen, 2006, p. 87). Some actors are more visible and others may be “hidden,” working behind the scenes to influence policy decisions (Kingdon, 1995). In our study, we considered all the actors directly and indirectly involved in the decision-making, their key interests and values, and the resources that might allocate them more or less power in this process—including their organizational position and various forms of capital (e.g., social networks, intellectual skills/expertise, and finances). District or charter leaders held considerable formal power, but so too did union leaders and state and local public elected officials.

• In the decision-making arena, actors will try to exert influence over policy decisions in overt and covert ways, including top-down authority, persuasion, information control, and collaboration (Fowler, 2009; Malen & Cochran, 2008; Marsh, 2016). In the case of school reopening, many decisions were made behind closed doors (e.g., labor-management meetings) and thus influence strategies were not always visible but could be gleaned from self-reports and media (e.g., efforts to promote the legitimacy of one's position on reopening by appealing to “science”).

• Finally, context matters. The sociocultural and institutional values, norms, and culture “mediate the power, preferences, and incentives of the players and shape the adoption and implementation of education policies” (Malen, 2006, p. 89). These contextual conditions shape not only who has power but also how actors think and behave and the political negotiations, dictating (consciously or unconsciously) what is and is not appropriate or acceptable in their local organization, community, and institutional context (Marsh, 2012). In the study of reopening, context could include federal, state, and local policies (e.g., delegating authority over school reopening); state and local partisan politics; governance structures of schools (e.g., level of centralization); and economic conditions. The school sector provided another important context that could potentially shape the dynamics of reopening decisions: in theory, the autonomy of charter schools could privilege different actors than those in TPS, such as allocating less power to teachers and ensuring leaders are more responsive to family preferences.

Past education policy research has applied a micropolitical lens to a number of school organization efforts and instructional policies (e.g., Ball, 1987; Galey-Horn & Woulfin, 2021; Grissom et al., 2015; Malen, 2006). Here, we extend this work and apply the micropolitical framework to help understand policy decisions of reopening schools during the early pandemic.

Methodology

We used data from a multiple, embedded case study (Yin, 2013) of five urban districts—Denver, Detroit, New Orleans, Portland, ² and Washington, D.C. (D.C.)—from March 2020 to July 2021. The states and four of the districts included in this study had participated in research we had been conducting prior to the pandemic, as part of the National Center for Research on Education Access and Choice (REACH) to understand efforts to improve access and equity in school choice policies. In the sections that follow we detail the case study sample, as well as our data collection and analyses.

Case Selection

Before the pandemic, we had selected five states (Colorado, Florida, Louisiana, Michigan, and Oregon) to represent variation in choice policies and settings, including geography, population, types of choice policies, and the maturity of these policies. In March 2020, this research agenda shifted to examine organizational and political responses to the pandemic across levels and sectors. Within each state, we sought to include one large urban district. We ultimately included Denver, Detroit, New Orleans, and Portland but were not able to recruit a Florida district. We also included D.C., which operates as a hybrid state district.

The five case districts provide variation in governance structure and choice context (Table 1). Portland has only few charter schools; Denver, Detroit, and D.C. have sizable charter populations; and New Orleans includes only charter schools. Each of our cases includes TPS (except for New Orleans) and charter schools, but with different governance arrangements. Denver operates under a “centralized portfolio model” (Bulkley et al., 2020), wherein charter schools have autonomy but are authorized and overseen by a more centralized and involved public school district. Detroit is a more fragmented governance landscape, with many different charter school authorizers, districts, and management companies, and no unifying governance or enrollment system. New Orleans is more of a “managed market” (Bulkley et al., 2020). All schools are autonomous charters authorized by the Orleans Parish School Board, and New Orleans Public Schools (NOLA-PS) serves as a central governance institution; and schools participate in a centralized enrollment system and some common regulations. Portland represents a more typical school district, in which the local school board has authorized a small number of charter schools that operate independently (e.g., no centralized enrollment systems, limited support, and communication from the central office). Finally, D.C. operates a parallel system: a TPS district and a large charter sector authorized by a separate charter school board, both under the oversight of the mayor, with little integration or interaction across the sectors. These varied contexts provide an opportunity to explore the micropolitical interactions of actors, policies, and contexts shaping reopening decisions.

Table 1.

Geographic, Demographic, Political, and Governmental Characteristics of Case Sites, 2020–21.

Case site	State	Student racial demographics	Political party in leadership	Biden vote share 2020	% Charter schools	School governance	Charter school authorizers
Denver	Colorado	Black: 13% Hispanic: 54% White: 24% Asian: 4% Other race: 5%	Mayor: D Governor: D State legislature: D	80%	26%	Portfolio management	Denver Public Schools
Detroit	Michigan	Black: 81% Hispanic: 13% White: 4% Asian: 1% Other race: 1%	Mayor: D Governor: D State legislature: R	94%	40%	Fragmented governance	Detroit Public Schools Community District, Wayne RESA, community colleges, universities
New Orleans	Louisiana	Black: 76% Hispanic: 11% White: 10% Asian: 2% Other race: 2%	Mayor: D Governor: D State legislature: R	83%	97%	Managed market	Orleans Parish School Board
Portland	Oregon	Black: 10% Hispanic: 20% White: 48% Asian: 12% Other race: 12%	Mayor: D Governor: D State legislature: D	84%	8%	District-authorized charters	Portland Public Schools
Washington, D.C.	—	Black: 65% Hispanic: 17% White: 13% Asian: 2% Other race: 3%	Mayor: D Governor: D Council: D	92%	50%	Parallel sector governance	D.C. Public Charter School Board

Political party in leadership data come from Ballotpedia (ballotpedia.org). D and R identify the elected official's political party or the party with majority control in the legislature; D represents Democrat, R represents Republican. Biden vote share in 2020 data come from the Harvard Dataverse (dataverse.harvard.edu/dataverse/electionscience). Precinct data are aggregated to the city-level. Student racial demographics and the percent of schools in the city that are charter schools come from the Common Core of Data via Urban Institute (educationdata.urban.org). Authorizers column lists actual authorizers, not possible authorizers (e.g., in Louisiana, the state board of education is also allowed to authorize charters).

Data Sources and Collection

The research team conducted 56 interviews via Zoom with central office administrators for TPS district, charter districts and charter management organizations (CMO); system leaders (e.g., superintendents); school leaders; teachers’ union leaders; and community and advocacy leaders, in the five sites between December 2020 and July 2021 (Table 2). In all districts, we selected leaders involved in TPS as well as charters. To assist with comparisons, we selected only principals from elementary and middle schools and prioritized schools with high proportions of low-income and racially minoritized students. Our interview protocols asked leaders to describe system-level or school responses to the pandemic, supports and services provided in a range of categories (e.g., academics, personnel, accountability, health and safety, and social and emotional support), and the role of other actors (i.e., state, local). For this paper, we draw heavily on interview questions about decision-making academics and learning, including school closure and reopening timelines; what informed these decisions; the nature of instruction and technology use; and challenges they faced. Interviews averaged one hour, and all interviews were audiotaped and transcribed.

Table 2.

Interview Data by Type.

Respondent type	Number of interviews (N = 56)
System Leaders (e.g., superintendents)	8
Central Office Administrators a	19
School Leaders and Principals	15
School Board Members	2
Teachers Union Leaders	3
Community-Based Organizations and Advocacy Leaders b	9

N = 6 in DC, N = 17 in Denver, N = 11 in Detroit, N = 12 in New Orleans, and N = 10 in Portland.

^aCentral office administrators included interviewees overseeing academics, operations, enrollment, research & accountability, charter authorization, communications, support & improvement, equity & diversity, and student health.

^bCommunity and advocacy leaders included those from organizations actively supporting school choice options, and broader community-based organizations serving families and children.

To triangulate and contextualize interview data, we consulted state and district/school system websites (checked regularly from the start of the pandemic), publicly available documents (e.g., reopening plans, executive orders), and data (e.g., on virus transmission rates, reopening dates by sector), and accounts from local newspapers, media, and social media (e.g., Twitter, Facebook).

We also use interview data from a subsample of parents and guardians in each of our sites. ³ Researchers completed a total of 46 parent interviews across the five districts (11 in Denver, 11 in Detroit, 5 in DC, 7 in New Orleans, and 12 in Portland). We draw from parents’ responses to questions about their experiences with policies adopted during the pandemic and views about their school leaders’ response to the pandemic, to strengthen our understanding of the context in which school and school system actors made reopening decisions.

Data Analysis

In our first phase of analysis, we uploaded and coded case study interview transcripts in Dedoose using an initial list of deductive codes derived from the larger study design and literature on crisis response, including codes for local context (e.g., sector, politics), pandemic effects and response (e.g., instructional mode, technology, health, and safety), vulnerable and historically marginalized populations (e.g., English learners, students in poverty, and students of color). Team members reviewed coded data to write detailed case profiles of each district (Maxwell, 2013; Saldaña, 2015), ranging from 60 to 130 single-spaced pages, seeking to capture the dynamics around pandemic responses, as well as broader issues of local context and politics. Team members followed a similar process for analyzing parent interview data.

In our second phase, we used these district profiles, parent interview memos, and coded data to conduct detailed single- and cross-case analyses via memoranda and matrices (Miles et al., 2014), to identify key dimensions of the micropolitics of reopening and emergent themes. For each case and across cases, we examined the full range of actors (e.g., governor, state legislators, public health officials, superintendents and other system leaders, school boards, unions, parent groups, central office administrators, regional leaders; see Table 3). We considered their preferences and values, sources of power and influence, how they exercised or lost power, sectoral differences and similarities, ⁴ and contextual factors affecting their exercises of power. We also constructed detailed timelines of the state and local actions taken to support learning in each district over the 16 months (Table 4), along with records of reopening statuses and COVID-19 transmission rates over time (Appendix B & C). Coauthors probed for similarities and differences within and across the districts and the conditions likely shaping the observed patterns. These memoranda and matrices provided the foundation for the findings presented here.

Table 3.

Major Actors Involved in School Reopening.

	Denver	Detroit	New Orleans	Portland	Washington, D.C.
School and district leader(s)	DPS centralized decision-making power in the district and drove decision modality decisions; decentralized decision-making on instructional planning to school level. Superintendent coordinated with other districts regionally.	DPSCD superintendent conceded some control over reopening through negotiation with teachers’ union; established “learning centers” policy. Some CMOs provided discretion to charter school leaders over reopening decisions.	NOLA-PS leadership centralized power to establish health and safety guidelines and phased reopening plans. NOLA-PS then allowed school-level decisions making within those parameters.	Superintendent decided schools would remain in remote learning relatively early in the Summer 2020 before school. The superintendent relied on the governor's executive order in March 2021 to reopen schools.	Superintendent works in concert with the mayor and public health officials to develop a reopening plan. Charter leaders are less publicly visible in their decision-making.
School board(s)	Placed emphasis on teacher and family safety, in part related to recent election of pro-union officials.	DPSCD school board supported and to an extent deferred to the superintendent.	Schools have separate boards from each other, so any board's influence was limited to their set of schools.	School board did not particularly exercise power or influence over decisions, mostly deferred to other decision-makers (e.g., superintendent).	Public and charter school boards voiced concerns of constituents but decisions remained with the mayor (public) or district and school leaders (charter).
Public health officials	District leaders designed plans in reference to local health authority recommendations. Local authorities issued their own recommendations, distinct from state and federal guidance.	District and school leaders designed plans in accordance with state guidance and with strong deference to health risks. Unions used guidelines in arguments about health risk.	District coordinated directly with city health officials and consulted guidance from national authorities to make decisions. District officials updated COVID-19 case threshold for closing schools over time in reference to new data.	Public health guidance greatly influenced initial decisions of the governor and superintendent; relatively strict COVID-19 case rate thresholds for reopening. Public health guidelines are relaxed to “advisory” over time.	City public health department was highly involved in planning for school reopening. Prioritized limiting the spread of COVID-19.
Teachers and teachers’ union	Prioritized health and safety concerns; did not successfully negotiate a COVID-19 case threshold for closing schools. District and charters felt concerned about teachers quitting.	Gained leverage through strike authorization during negotiations; secured teacher choice over modality and COVID-19 case threshold for closing schools. District and charters concerned about teachers quitting.	Teachers preferred staying remote for longer but without a sector-wide union had little collective power to influence reopening decisions.	Teachers’ union pushed for health and safety measures and against in-person instruction. District perceived the union to have significant influence, but they had little leverage in reopening negotiations in Spring 2021 after the governor's executive order.	DCPS teachers’ union won significant health guidelines but had to concede on teachers returning in-person for quarter 3. Charters concerned about teacher well-being and threat of union organizing in the sector.
Parents	Parents in both sectors expressed concern for health and safety issues and support for school health measures, though frustrated with communication about them. They weren’t satisfied with instruction but were understanding about schools’ efforts.	Parents in both sectors expressed concern for health and safety issues and support for school health measures, though some frustration with communication about them. They were also understanding about the challenge schools faced.	Parents expressed concern about both health and safety issues and academic and socioemotional consequences of the pandemic. Support for school health measures. Modality choices were driven by levels of comfort with health and safety.	Parents felt concerned about health and safety, and unsatisfied but understanding about remote learning. The superintendent said that parents did not influence decision-making. District perceived white and affluent parents to prefer reopening.	Parents were concerned about their children's health and safety, and most expressed satisfaction with or appreciation for school measures. They weren’t satisfied with instruction but were understanding about schools’ and especially teachers’ efforts.
Governor and mayor (context)	Governor deferred reopening authority to individual districts and offered guidelines.	Governor established health and safety guidelines and reopening recommendations in line with those; allowed for district-level decisions about reopening until push for all schools to offer an in-person option by March 2021.	Governor deferred to local authorities on reopening. The mayor coordinated with private-sector leaders to implement public health measures explicitly aimed at reduced community spread and keeping schools open.	Governor orders shaped reopening timelines. Governor took a cautious approach to the pandemic, with most Oregon schools remaining closed through most of the year based on state guidance. State prioritized teacher vaccination in December 2020. In March 2021, the governor executive-ordered schools to reopen.	Mayor exercised strong authority over the DCPS reopening plan, in collaboration with the DCPS superintendent and city public health officials. Prioritized building trust and establishing legitimacy for eventual school reopening.

Table 4.

Description of School Reopening Timelines.

Case site	Early response Spring/Summer 2020	Preparation for 2020–21 school year Summer/Fall 2020	Start of the 2020–21 school year Fall 2020	Changes over the 2020–21 school year Fall 2020–Summer 2021
Denver	Temporary school closures in March followed by building closure and remote learning through the end of the school year per Governor's orders. Schools focus on distributing food, technology, and other materials for remote learning; and prioritize checking on student and family well-being.	State government releases guidance for reopening in July. DPS plans reopening and negotiates with teachers’ union; union fails to secure a case-rate threshold to close schools. DPS and charters, somewhat collaboratively, decided to make school closing decisions in response to county health data. DPS also coordinated reopening standards with surrounding districts.	At the beginning of the school year the overwhelming majority of districts in the state were virtual or hybrid. Families had until mid- or late-September to decide if children would remain fully virtual or not. By late October, Denver completed a phased return for all elementary students being back in-person.	In November, case rates spiked across the state. DPS schools switched to virtual; some charters remained hybrid. DPS and (many) charter schools start to phase in-person in January, with most students in-person by spring 2021. Vaccines became eligible for school staff in January, broadly for adults in March; extended vaccination sites and incentives for teens to be vaccinated in May.
Detroit	Governor orders schools closed temporarily in March, followed by building closures for the remained of the year. starting 3/16, initially for a couple weeks but eventually for the remainder of the school year. Schools focus on distributing food, technology, and other materials for remote learning; and prioritize checking on student and family well-being. The state required schools to submit distance learning plans.	Governor requires a “Return to School” plan from all districts and establishes general public health guidance based on case-rates. In-process negotiations between DPSCD and the teacher's union shift from collective bargaining agreement to memorandum of agreement related to the pandemic. In the process, the union authorized a strike. DPSCD agreed to allow teachers to choose in-person or remote teaching, and set a 5% COVID-19 case-positivity threshold for closing schools.	The state allows school districts to open based on their approved “return to school” plans. The large majority of DPSCD teachers choose to teach remotely; DPSCD establishes “learning centers” in schools, where noninstructional staff supervise and assist students in the building as they participate in remote learning. Charter schools mostly offer remote or hybrid options.	In December, rising COVID-19 cases lead DPSCD and most charters to return to remote-only. Most schools return to offering an in-person option by the spring. In January, governor urges MI districts to offer an in-person option by March 1 (within Biden's first 100 days). School staff become eligible for the vaccine in January.
New Orleans	Governor announces temporary school closures in March, and again in April; announces closures for the rest of the school year in mid-April. Charter schools coordinated centrally with NOLA-PS to deliver or provide food. The district provided technology in March, then left responsibility to schools in April onward.	In July, the district released a “Roadmap to Reopening” with guidance from local and national public health officials. The plan tied reopening to public health phases; and the all-charter district established a unified school calendar for the first time. Within this guidance and calendar, schools have autonomy. In August, NOLA-PS announced a goal to provide in-person learning by early September; parents still have the option to remain virtual all year. A city-wide group of public and private leaders collaborated around reopening, supporting more stringent public health measures (e.g., mask mandates, curfews on bars, limited indoor seating) as necessary measures to facilitate in-person learning.	Charter schools begin the school year in distance learning; private schools mostly return to in-person learning. Reopening was further delayed due to hurricanes. Phased in-person reopening (elementary, then middle, then high school) took place throughout September and early October.	In September, a Court of Appeals ruled Louisiana's charter schools must recognize and bargain with unions if that is the desire of teachers and school employees. Some union organizing proceeds; unionization remained very low. In January, schools switched to virtual learning for three weeks, following an increase in COVID-19 cases. Hybrid and in-person incrementally reoffered to elementary and middle school students; hybrid or remote (no full in-person) offered to high school students through May. Vaccination for school staff starts in February.
Portland	Governor announced temporary school closures in March, followed by closures for the remainder of the year in April. PPS offers comprehensive distance learning options, and works with community organizations to provide internet and technology.	In June, the state department of education released “Ready Schools, Safe Learners” guidance on reopening; required reopening plans for approval in August. The state also announced health metrics to guide reopening decisions in late July (and updated in early August). PPS developed in-person plans for the fall but in July the superintendent announced in July that schools would start the year closed.	PPS started the year fully remote as did the few charter schools in the city. The teacher's union took a strong stance against reopening, citing concerns about health and safety such as poor ventilation in school buildings. The superintendent viewed union strength and teacher concerns as an obstacle to reopening, appealing to the governor to issue an executive order for school reopening, which would neutralize union leverage.	In October, the state relaxed some health guidance; and in January, guidance metrics became advisory rather than compulsory. In December, teachers received priority status for vaccines. In March, the governor ordered schools to provide in-person instruction for K-5. For reopening, the union negotiated that teachers would not be required to conduct both in-person and virtual learning, and secured some additional health and safety measures. Families were still able to choose online or hybrid.
Washington, D.C.	DCPS temporarily closes some schools in March, followed by the mayor temporarily closing all schools the following day. In April, school closures are extended for the remainder of the school year. Charter schools similarly close. Schools, supported by several city offices, focus on communicating with families and providing food and technology. The mayor's office arranges discussion between DCPS and charter district leaders and local public health officials.	Mayor's office, superintendent's office, and department of health officials develop a plan for remote learning for the first quarter and probably the second quarter. Charters were required to submit “Continuous Education and School Recovery Plans” to the superintendent's office. Most charters plan for remote start. In-process negotiations between DCPS and the teachers’ union shift from collective bargaining agreement to memorandum of agreement (MOA) related to the pandemic. Negotiations continue until MOA is signed in December 2020	DCPS starts with remote learning, but offers “CARE classrooms” to accommodate students with special educational needs and children of essential workers. Charters similarly start remote, some offer small in-person options similar to “CARE classrooms.”	DCPS and union finalize MOA, which includes health and safety provisions but no case-rate thresholds for reclosing schools. DCPS schools reopen in February (quarter three) in accordance with the MOA. Parents can choose in-person or virtual. Some rank-and-file union members push for stricter public health requirements but no changes are made. DCPS conducts safety and health walkthrough audits to boost family confidence in returning in-person. Many charter schools offer in-person options in the second half of the school year; many also remain remote-only.

Findings

Overview of Reopening Processes and Timelines

In all five school systems we studied, governors (or the mayor in D.C.) ordered schools to close in spring 2020, first temporarily and eventually for the remainder of the 2019–20 school year. Over the summer and leading into the fall, school and system leaders prepared for the 2020–21 school year. For all of our cases, the decision-making period before the start of the school year involved the preparation of formal plans for reopening during the pandemic, which addressed policies and procedures for instruction, health, and safety. School and system leaders responded to requirements or guidelines from state and local lawmakers and public health officials and engaged with other actors as they prepared reopening plans. Across cases, district administrators tended to drive reopening planning, with school boards supporting their efforts rather than leading (Marsh et al., 2022). Across our case sites, planning involved alignment with local or state health guidance and varying degrees of centralization and cross-sector coordination. In most sites, teachers were central to reopening planning, including TPS districts establishing memoranda of agreement with teachers’ unions, and charter schools actively engaging staff about their preferences and concerns related to health measures and instructional modality. TPS and charters also considered family preferences, mostly relying on surveys and other forms of public engagement to solicit feedback.

The school reopening timelines for these five cities reveal striking similarities. For each of our cases (except Portland), the period of decision-making prior to the 2020–21 school year was pivotal to establishing the trajectory of school reopenings: TPS districts followed the parameters set forth in their initial plans. The available data on school modality also show similar reopening patterns by sector for each case site:

• TPS and charters in our case sites mostly started the year fully remote, with in-person options phasing in (either shortly or later on) over time, and with schools mostly closing again during the winter spike in COVID-19 case rates.

• Reopening timelines varied among charters in each site, but they broadly followed the same patterns as the TPS in their city.

• In each site, all schools also prioritized giving parents options when reopening: they tried to make in-person options available, but even when they pushed most strongly for reopening they did not take away virtual options.

Still, we saw important differences in when and how schools reopened across our case sites.

• New Orleans schools opened the earliest, with in-person options available at most schools by October, punctuated by closures during a winter spike in COVID-19 cases.

• Denver leaders, who planned to phase-in reopening in the fall, delayed full reopening until the winter, after a decline in COVID-19 cases.

• Other than some initial in-person options for children of essential workers and children with special educational needs, Washington, D.C., opened its schools after the winter COVID-19 surge. This was consistent with its initial plan.

• Portland, with the consistently lowest COVID-19 transmission rate among the cases, remained remote-only the longest of any of our case sites, only reopening its schools in the spring after an executive order from the governor.

• In Detroit, the TPS district opened its buildings immediately as “learning centers” but offered little traditional in-person instruction until later in the school year because teachers could choose between remote and in-person instruction and many opted to stay remote; and the majority of charters did not reopen until after the winter COVID-19 surge.

What explains these patterns? In what follows, we describe how similarities and differences in the interactions among actors and the contexts for decision-making shaped school reopening (Figure 1).

Figure 1.

Framework for the Micropolitics of School Reopening in case sites.

The Micropolitics of School Reopening

Concerns About COVID-19 and Strategic Adherence to Public Health Guidance

Public health concerns were salient and central to plans for reopening all of our case sites. We found evidence that district leaders appealed to and embraced public health guidance strategically: it aligned with their values and it helped them establish legitimacy and minimize risk in a politicized public health climate.

Adherence to public health guidance served to legitimate decision-making in light of the local parents’ and teachers’ uncertainty and fear of COVID-19. School and district leaders held concerns about COVID-19 illness and death as a result of school reopenings alongside the academic and socioemotional consequences of schools remaining closed. This was especially the case in cities that had previously experienced high levels of sickness and death earlier on such as New Orleans and Detroit (Appendix C). One system-level leader in New Orleans put it this way:

It was grueling, even as I just think to even say this to you, in the very beginning, it's almost like every single day, it's like if I hear that one more person passed away in New Orleans I'm just going to scream…[Our roadmap for reopening] was created to say, like, “Hey, safety's going to be the priority, we lost too many people. Our students who are attending our school, that was their family members that are deceased.” Just the number of Black individuals that died, it was just unbelievable in that short period of time compared to other districts in the state of Louisiana.

The quote demonstrates the emotional toll of COVID-19, both for students and their families and for the leaders charged with planning for school reopenings in light of those circumstances. In this example and others, school and district leaders viewed public health measures as important for establishing legitimacy in their own districts, and as distinct from the contexts of other districts.

School and school system leaders thus described strategically adhering to public health guidance to legitimate their decision-making in the context of public concern for health and safety. For example, a district leader in Portland stated: “Well, I don’t need the community thinking that I am unilaterally deciding what's safe or what's healthy. I need professionals and subject matter experts to tell us, ‘These are the guidelines.’” Others we interviewed echoed this appeal to expertise—for example, stressing the importance of following “the science” and emphasizing specific recommendations they adopted from the Centers for Disease Control (CDC). In part, adhering to public health guidance helped school system leaders buffer against risks and criticism from different interests. As one public health official in Denver explained:

There were lots of pressures the superintendents had to respond to from parents, students, unions, etc. The superintendents just told us again and again how helpful it was to have a uniform set of guidance to say, “Look, we’re following the Metro Denver Partnership for Health guidance, get off our backs.”

In all of our cases, the strategic use of public health guidance to legitimate decisions meant paying close attention to advice from local and national health agencies and using resources to implement a range of health and safety standards.

School and system leaders did not only appeal to public health guidance to support keeping schools closed; they also did so as they addressed hesitance or opposition to reopening, in particular from teachers and teachers’ unions. In expressing their positions on reopening (described in the next section below), teachers and unions appealed to the same public health guidance that school and district leaders turned to for legitimacy. As districts formalized agreements with unions and responded to teachers’ concerns, they asserted that their own positions were aligned with public health authorities. For example, a board member in Denver explained how they resisted defining a specific case-rate threshold for closing schools: “Labor wanted to not go back to school until community spread was below 7%. We didn't want to commit to a hard number like that…our public health officials didn't suggest that.” Broadly accepted support for public health caution provided teachers and unions with favorable conditions to promote increased safety measures. Thus, school and school system leaders strategically navigated their embrace and interpretation of public health guidance to support their own interests.

Labor-Management Relations and Perceptions of the Teacher Labor Market

Equally salient as public health guidance was to reopening decisions was the role of labor dynamics, including labor-management relations, political action from teachers’ unions, and leaders’ perceptions of the teacher labor market. TPS districts in each site collaborated or entered into formal discussions with teachers’ unions to establish health and safety conditions and instructional expectations. Common concern for teacher well-being and teacher shortages, and differences in union leverage and influence, highlight similarities and differences among our cases.

School and system leaders’ perceptions of teacher well-being and the teacher labor market strongly influenced decision-making. Across all of our cases, school and district leaders acknowledged the hardship teachers faced during the pandemic and placed a great level of emphasis on supporting their well-being. In many cases, leaders also expressed concerns that teachers might quit or retire during the school year, leaving them with staffing shortages. This may have given teachers’ unions additional leverage, but it also influenced leaders in nonunionized settings (i.e., charters). In Detroit, for example, one charter school leader explained her sensitivity to the teacher well-being and labor market conditions this way:

And I hate to say it like this, but I don’t know any other way to say it, but I’m not riding teachers as much for, like, academic outcomes, because we’re in a pandemic…I mean, we’re still holding them accountable, because I am very much so like this. I’ve been just trying to leave people alone, to be honest with you, because I don’t want them to quit. Because the fact of the matter is having a teacher, like what are we going to do? Like I have to work with a sub who is not really with it, and then our kids suffer. I’ve eased off of it. And not even just the academic component, but just some of the things I was super hard on before.

As the quote suggests, support for teacher well-being and concerns about burnout and teacher vacancies influenced decision-making across sectors. These perceptions of teacher labor market conditions made school and system leaders more sensitive to and considerate of teacher concerns when planning for reopening.

In this context, teachers’ unions sought to influence reopening. Unions consistently prioritized the health and safety of their members, which they also connected to the health and safety of students and families. A representative from the Detroit teachers’ union described their position this way:

We can’t compromise health and safety. That's a non-negotiable. I think that for us, we wanted to make sure that we offer some level of face-to-face instruction for those who are comfortable. We wanted to make sure that was safe, safely done. We wanted to sing from the same hymnal, so if it's going to be CDC guidelines, let's stick to CDC guidelines. If it's going to be mitigation of strategies that include but aren’t limited to PPE, face masks, six feet of distancing, hand washing, proper ventilation, let's get to the same guidelines here…One of the other things that we wanted was we needed a trigger, a number that says we'll continue face-to-face if it's this, as long as it's this, we’ll stop it if it becomes this.

In addition to highlighting union priorities and common points of negotiation around health and safety, the quote hints at tactics that were common across sites. Unions and their teachers described a high level of risk, pointing to health hazards such as high rates of community COVID-19 spread, poor condition of building ventilation, lack of capacity for social distancing, and the need for adequate protective equipment. They also made direct appeals to official health and safety guidance, as seen above. Unions used these tactics when negotiating with TPS districts and making broader appeals to the public. Our data, however, indicate different tactics by unions, and different responses by districts.

Compared to the other cases, Detroit's teachers’ union took the strongest action during negotiations and substantially influenced both the timing and nature of reopening. The rank-and-file members in Detroit's union authorized a strike if the district did not agree to acceptable health and safety terms. After a highly publicized battle, the district and union signed an agreement before the start of the school year. Unlike any of our other cases, the union secured teachers’ ability to choose whether they would teach in-person or remotely for the entire year. Ultimately, 95% of teachers started teaching remotely. Thus, even while the district opened “learning centers” for students at the start of the school year, its concessions to the union severely limited its ability to offer in-person instruction. The district also agreed to a 5% COVID-19 case-positivity threshold for closing schools, which came into effect during the winter surge.

The Portland and DC teachers’ unions also applied significant pressure throughout the school year, influencing school reopening. In Portland, the union argued that building conditions (e.g., poor ventilation) made reopening unsafe; and, especially when the district announced its reopening plan, they echoed concerns from Portland families of color about the health risks of returning to school in light of COVID-19's racially disproportionate impact in the city so far (Campuzano, 2021). After the governor issued an executive order to reopen schools (see “The Role of State Government”), the union had little leverage to influence the timing of reopening; but they were able to secure some health and instructional guarantees, for example, that teachers would teach either in-person or virtually but not both (i.e., hybrid).

In DC, the union's collective bargaining agreement had been scheduled to end with the 2019–20 school year, and negotiations for a new CBA were underway when COVID hit in March. Negotiations continued but shifted focus to the development of an extensive Memorandum of Agreement on what existing CBA provisions would have to be suspended in response to COVID-19. Negotiations continued for the rest of the calendar year, broadening to include requirements for health and safety provisions at school sites, public information about assessments and conditions at those sites, and opportunities for teachers, parents, and community members to walk through each site in advance of its reopening. Provisions did not include fixed targets for COVID-19 case rates or a timeline for reopening. The parties finalized and signed the MOA in December 2020, and it strongly shaped conditions when schools finally began to reopen in February.

In Denver, teachers’ unions also appeared to play a less influential role in the timeline of reopening, though they also influenced the nature of reopening in some ways. The district coordinated its reopening plan with other districts in the region, which, as one district leader explained, made it “harder for the associations [to oppose reopening]…if the whole group makes decisions and then sticks to it.” The superintendent also formally invited union leadership into its centralized, collaborative decision-making body. The Denver union expressed interest in requiring the district to set a defined COVID-19 case threshold for closing schools that they had wanted, which the district refused to do. Denver school leaders nevertheless delayed reopening until a period of lower case rates, underscoring the role of other factors beyond union influence in shaping reopening timelines.

Finally, New Orleans, with almost no unionized teachers, had little union influence. As one New Orleans school system leader put it:

Union members or leaders actually attended our school board meeting and they sent letters. There were many places where we were aligned, there were some places where we were not. Many times the unions locally are, I think, using that national voice and whatever those talking points are. You just take that for whatever it is and keep moving.”

Consequently, teachers may have had less organized power to influence reopening decisions. For example, one community leader in the city remarked:

A lot of teachers know, and a lot of principals know that people get fired for petty things…and so, we hear a lot from folks who are like, “Hey, I want to be a teacher and I want to support students. I want to support low-income families, and this is why I'm here. But also, the politics of where I'm doing this work prevents [me] from being so loud and out about it.”

This limited influence of teachers and unions stands out as one of two major contrasts between New Orleans and our other sites (see “Other Governance and Institutional Factors”).

Parents as Cautious and Individual Actors

Parents tended toward deference and individual action around school reopening rather than through concerted and collective demands. School and system leaders used formal but limited engagement strategies, such as surveys, to quickly and efficiently obtain parent input to inform decision-making. Ultimately, leaders interviewed believed their plans were responsive to parents’ varied needs and did not perceive a critical mass of opinion or collective demands from parents for a different reopening timeline. A general posture of caution and deference among most parents informed decision-making; and the absence of clear and organized interest from parents allowed other interests and actions (discussed in this paper) to influence school reopening as they did.

Before and during the school year, parents mostly engaged with schools across our sites as individual rather than collective actors, which dampened the demands school and system leaders felt from parents. There were some exceptions: in Portland and Denver there were a few organized efforts by some parents to call for reopening schools; and in some instances, parents contributed to district- or school-based committees. But, for the most part, parents shared their input with schools and districts in response to formal outreach through surveys and virtual town halls. For example, one parent in New Orleans recalled the principal of his children's school contacting him directly encouraging him to “do a little survey, about would you like your child to do remote learning or virtual learning” so that the school could “make their decisions” on modality choices. School and system leaders recognized surveys and other formal forms of communication aligned with more of an individual mode of engagement with parents than a collective one. As one principal in New Orleans put it, “There's not this group that [parents’] voice can be sent through and amplified, and so they're just individuals. So we lose their perspective, I think, when we're making decisions.”

Parents interviewed ⁵ often held cautious views on reopening themselves and were typically supportive of how schools were handling it. The novelty and instability of the pandemic created new challenges for parents, which left many feeling ambivalent about their school's response; yet, they often viewed schools as trying their best, given the circumstances (Tong et al., 2022). This ambivalent-but-understanding attitude is captured in the following quotes from a New Orleans parent and a Portland parent:

Who knew how to respond to that? I mean, the damn president didn’t even know how to react to it…I mean, I can’t say that they did wrong. They did the right thing. I mean, at that point in time, you know, everybody's fearing for their life. So, sending everybody home and shutting down everything could have been all right. (New Orleans parent)

I understand that different areas have different responses, you know, based on what's happening…So, you know, I’m, I’m a lot more forgiving on that because I understand that their job is very difficult, you know, exactly what do you do?…Then I found out a little bit more too, that they have to collaborate, you know, with like…county officials, state, you know? There's a lot of communication [in] the pipeline that has to happen for them to make these actionable choices. (Portland parent)

As both quotes illustrate, parents interviewed were mostly gracious about the decisions made by their children's schools. This understandings meant that schools felt less pressure from parents around reopening decisions.

Central to the influence of parents on reopening was the racialized and class-based way that school and system leaders perceived and responded to parent preferences. First, leaders’ imperative to balance educational goals with health and safety was racialized. On one hand, they understood the stakes for reopening schools in terms of racial equity (though sometimes cloaked in race-evasive terms). For example, a district leader of the Detroit TPS said their main concern with reopening was “equitable access to education…if suburban school districts were going to be opened in the Fall, then [our district] has to be opened.” In other words, it would be inequitable to keep schools closed for a district of primarily low-income students of color when neighboring white or more affluent students were receiving such opportunities. On the other hand, having witnessed greater levels of illness and death among their cities’ nonwhite residents, they were concerned about racial disparities and the health impacts of COVID-19; thus, they were sensitive to the risk that the spread of COVID-19 through in-person schooling might pose for students of color and their families (discussed above).

Second, school and system leaders saw preferences for reopening in racialized (and class-based) terms, in particular, perceiving white (and affluent) families as most strongly desiring in-person instruction. They formed these impressions from the interactions they had with families with different racial and class backgrounds and their observations of demographic differences for students attending school in-person and remotely (when options were available). In New Orleans, for example, a district leader explained that white and affluent students tended to most often opt for in-person, whereas for “students of color, we see more distance learning.” Likewise, a Portland district leader explained that he had “to look out for every student” and that “there's differences of opinion,” noting that the parents most vocally calling for school reopening “tend[ed] to be more [from] white, affluent neighborhoods.” Thus, school and system leaders came to believe nonwhite families were not strongly demanding in-person learning, especially against the backdrop of COVID-19's racially unequal health impacts. ⁶ Especially because school and system leaders understood reopening as an issue of racial (and socioeconomic) equity, their perceptions reinforced the feeling that cautious adherence to public health guidance was the best course of action.

More than anything, school and system leaders in our case sites appeared to believe that families wanted a remote option and the ability to select this option when they felt more comfortable with it. As such, they appear to have rarely (if ever) eliminated remote learning options, even after in-person options became available. As one New Orleans district leader put it, “A distance learning option [is] always on the table for any individual.” Thus, parent preferences influenced school and district leaders in two ways: (1) their caution reinforced district leaders’ adherence to public health guidance, and (2) they did not disrupt the reopening plans that resulted from public health considerations, negotiations with unions, and other governance and contextual factors.

The Role of State Governments in Shaping Local Contexts

State governments, and in particular governors (or the city government and the mayor in D.C.), played another important role in the reopening across sites. Notably, the strong role of the Oregon state government helps explain why Portland schools remained remote-only for the longest of our case sites despite lower levels of COVID-19 case rates throughout the year. The mayor of D.C. and her Deputy Mayor for Education also exercised strong central authority to shape school reopening in that city. In our other three case districts, state governments provided districts with discretionary authority, and school and system leaders still exercised careful adherence to public health guidance.

Oregon's state government set Portland's schools on its path to remote-only instruction for most of the year and also played a decisive role in their eventual reopening. The governor of Oregon took a cautious approach to COVID-19, prioritizing strong health and safety measures. This high degree of caution in Oregon, which stands out given relatively lower case rates, may have been in response to early school-based spread of COVID-19. The first documented case of COVID-19 in Oregon involved a school district employee, increasing the salience of the health risk in schools from an early point (Gordon, 2020). The state's department of health developed metrics to guide school district reopening over the summer, and the state department of education required reopening plans aligned with these health metrics to be submitted for approval before the start of the school year. Most districts started with remote-only learning based on these metrics. The state eased some of its guidance in the fall that allowed schools in less densely populated areas to reopen, but most schools in densely populated areas like Portland remained closed. The state made two further shifts to ease reopening in the winter: teachers were placed in the first priority group for vaccination in December 2020, and the reopening metrics from the state became advisory rather than requirements in January 2021. Still, a large number of schools remained closed, until the governor intervened in March 2021 with an executive order.

The governor of Oregon's order required schools to provide a hybrid or fully in-person option for K-5 by March 29 and all other grades by April 19 or risk losing funding. In her order, the governor made an appeal to expertise similar to those that had previously been used to support caution: “the science and data is clear, schools can return to in-person instruction with a very low risk of COVID-19 transmission, particularly with a vaccinated workforce” (Brown, 2021, p. 3). This action enabled the Portland TPS to offer in-person options for the first time. As a TPS district leader explained, the order “orient[ed] everybody's mindset towards a reopening” and “certainly signaled to everybody, it was time to really work out and get down the list of what needed to happen in order for schools to reopen pretty quickly.” District leaders in Portland described the governor's intervention as necessary because of the local political conditions, and in particular teachers’ union resistance to reopening (see “Labor-Management Relations” above).

In Washington, D.C., the mayor exercised her significant authority over the school system to press for reopening in November 2020. The mayor and her Deputy Mayor for Education worked closely with public health authorities, placing a strong emphasis on establishing trust among parents in school health and safety protocols. They did this amid the school district's protracted negotiations with the teachers’ union on a Memorandum of Agreement that would set detailed conditions for reopening and amid public reluctance by many teachers and parents to return to in-person instruction. The district and the union signed a detailed MOA in December, but local COVID conditions and the complex and expensive logistics of preparing schools for safe reopening meant the schools would remain mostly closed until February 2021. Thus, the exercise of executive authority included adherence to public health guidance and resulted in a similar reopening timeline to that of Denver and Detroit, where school and system leaders had more discretion.

For Denver, Detroit, and New Orleans, the state governors mostly gave discretion to individual districts over their reopening plans. Michigan's governor provided reopening guidelines in line with state-defined phases of the pandemic and required reopening plans to be submitted, but ultimately gave discretion to districts. In contrast, Colorado's governor encouraged districts to consult guidelines from their local health authorities but did not require plans; and Louisiana provided little state guidance. Still, Denver and New Orleans developed guidelines in close adherence to recommendations from their local public health authorities. Although there are differences in the reopening timelines for Denver, Detroit, and New Orleans, their similarities highlight the importance of governors devolving decision-making authority to the local level. With their discretionary authority, school and system leaders in our case sites stayed closed for longer than many other districts in their states and returned to remote-only instruction during a winter COVID-19 surge when other districts did not.

Although we have highlighted the role of governors so far, state legislatures also played a role by passing funding and public health bills during the pandemic. Most states passed hold-harmless bills—tying current-year funding to prior-year attendance or enrollment, setting limits on revenue declines, or providing supplemental funding—to buoy school finances as enrollment tended to drop (Center for Public Education, 2021; Menefee-Libey et al., 2022). With shored-up budgets and additional money from federal legislation, districts prioritized purchasing personal protective equipment and other health and safety resources, and in some cases provided bonuses or hazard pay for teachers. For our case sites, states also offered teachers relatively early access to COVID-19 vaccines, which was seen as a step toward ensuring that teachers could safely participate in in-person instruction. Vaccine availability, however, tended to coincide with declining COVID-19 rates and planned reopening phases for our case sites, so it is unclear without additional cases how significantly they influenced reopening timelines.

School Sector

In each of our case sites, charter schools had at least some formal autonomy from the local TPS district (and in some cases complete autonomy) to make reopening decisions; yet, broadly, they tended to follow the same timeline as the TPS district. ⁷ Our data indicate that school sector conditions mattered less because charter schools faced similar political pressures as the TPS and tended to respond similarly. Like their counterparts, charter schools likely adhered to local public health guidance to avoid risk and establish legitimacy. They also faced similar teacher labor market conditions: even without formal pressure from organized unions, responding to teachers’ needs and concerns about labor shortages factored into charter leaders’ decision-making. This held true across our case sites, which each have different school governance structures. Taken together, charter schools’ responsiveness to political factors and family preference help explain why they followed similar reopening patterns to their local TPS districts and why many charter schools in our case sites stayed remote-only for even longer.

The charter school leaders whom we interviewed, like their TPS counterparts, mostly took a cautious approach to the health and safety risks of COVID-19. From school, district, and management organization leaders, we heard similar perceptions of public health concerns among staff and families, and similar language about following local and national public health guidelines. One CMO leader in Detroit put it this way:

There was a lot of fear. So we wanted to respect, where everyone was…Fear from teachers, staff, parents, students, so I just think there was, there still is, some fear with some, even if they’re vaccinated, there's still fear…So, you kind of have to balance that out but I think, [we were] really listening, following science, the guidelines from the CDC.

In addition to adhering to public health guidance, we heard evidence that some charter schools looked to TPS for cues for their decision-making. For example, a CMO leader in Denver described their reopening plan like this: “The plan basically is now follow the [authorizing] districts’ [plans] because they’re working with our public health authorities…the districts are deciding to bring back elementary next week, and we are going to do that too.” Another central officer administrator from the same CMO remarked:

At the beginning of the year, [our CMO] was setting its own metrics for when we’d be open or closed and it ended up being more conservative than the district…I actually think it's much better to leave giant safety decisions to the district leadership…they have more access to public health leaders and that's who should really be informing these decisions at this point.

As these quotes suggest, some charters saw the TPS as legitimate indicators of public health safety, and they may have believed that acting in concert with TPS and public health authorities would legitimize their decision-making to staff and families.

Charters also may have viewed alignment with TPS as important in managing the cooperative and competitive relationships with TPS districts. For example, a charter school administrator in Portland expressed concern “that if the charter schools can't offer the same options, they won't look like a comparable option anymore.” Likewise, a charter school leader in Denver (from a different CMO than the one cited above) recalled, “I remember someone saying ‘Y’all are going to open ahead of the district? You better not screw it up, because you have thousands and thousands of eyeballs on you.’”

In addition, charter schools were similarly sensitive to their teachers’ concerns and the state of the teacher labor market. In our case sites, almost no charter schools have unionized staff and thus did not engage in the kinds of labor negotiations that TPS districts did. Yet, charter school leaders were similarly concerned that teachers would quit or retire if asked to return in-person too quickly, leaving them with shortages this year and burdening them with vacancies to fill in the coming years. For example, one CMO leader in D.C. explained that “asking staff to come back and requiring staff to come back…it felt really risky. From, for us, from a [threat of] unionization perspective, from a staff morale perspective, also from just a culture perspective.” Another leader from the same CMO explained they were worried because “this year is by far our best retention year we’ve ever had…[but] I'm now worried about a cliff…Two years from now or a year from now, we're going to have 30% of teachers leave or something like that.” Therefore, charters were also responsive to teacher preferences, either gauging their comfort levels before making a reopening decision or allowing them to choose between remote and in-person instruction. New Orleans stood out in this respect, where teachers appeared to have a relatively smaller influence over school reopening decisions (discussed above).

In addition, charter families interviewed had similar health and safety concerns and attitudes toward reopening as TPS families interviewed (Tong et al., 2022). Given families’ health and safety concerns and interest in remote learning (discussed above), many charter schools in our case sites seem to have used their autonomy to remain remote-only in response to perceived family preferences.

Taken together, how do these political factors and family preferences explain the reopening decisions of charter schools in our case sites? The answer is well-represented by the following quote, from a charter school leader in Denver:

[Earlier in the year] I put the decision in our families’ and teachers’ hands, and our teacher said, “If we go remote, our kids will not learn for four weeks.” And so they decided to stay in person…[Our decision to switch back to remote learning] was driven by families starting to get nervous, teachers starting to get nervous. And at that point, when we decided to close, I think Denver County was up at 12.2%. And that's where you really have to start. And thinking back on this, and like, it's honoring voice. Like when teachers made the decision, “We’re going to come back sooner than the district,” we feel good about it. Denver County was at 2.8%. And so, in the words of one of my teachers, [they] are like, “We signed up for 2.8%, we did not sign up for a 12.2% [COVID-19] positivity rate in the county.”

Similar to leaders in TPS districts, charter school leaders were cautious about the health risks of COVID-19 and adhered to public health guidelines and were responsive to the concerns of their teachers and families. Not all charter schools made the same decisions. Some offered in-person options earlier or more often, and some remained remote-only longer or for the entire year. ⁸ On the whole, however, political factors and family preferences influenced public and charter schools alike and led to similar reopening timelines in both sectors.

Although common political and contextual influences led to similarities in when charters and TPS reopened, we have some evidence that sector governance and autonomy influenced how charters reopened. Autonomy allowed some charters to try different approaches to scheduling and delivering in-person instruction than TPS in their districts; and governance differences may have constrained or facilitated that autonomy. For example, in Denver, charters had a good deal of autonomy in their response to the pandemic. One charter school principal explained that, since “instructionally we had freedom,” the school was able to modify its schedule and staffing assignments to adjust the balance of live-instruction provided for students who remained remote. As part of the portfolio model, however, some Denver charters rely on the district for school-based transportation. When the district changed its bus schedule to accommodate social distancing rules and align with its new in-person instructional calendar, these charters had to adopt the same TPS schedule in order to continue providing transportation. Facing these constraints, one charter school principal explained, “I just feel we’re under the direction of the DPS central way more than we ever have been”:

…we’ve always set our own calendar year and schedule in terms of daily hours…Then the district had to come up with a very complicated busing plan. Basically they were like, “This is your bell time.”…We lost an hour and a half a day of instruction for our kids, which it's not what I would’ve wanted…I work at a school where a lot of kids ride the bus, so it's not it could be, “Oh, screw that, we'll just do what we want.” Those are kids who really need that transportation service.

The issue of school buses in Denver is just one example that illustrates how the governance context for charters could enable or restrain how they planned for reopening. New Orleans, where the district imposed the first unified school calendar since the schools became all-charter, offers another example of how centralized governance could restrict charter autonomy. In Detroit's fragmented system and D.C.'s parallel system, by contrast, charters did not face these kinds of constraints.

In sum, while the regulations and governance differed for charters, they existed in the same context as their TPS counterparts. Those influences—the same local COVID-19 rates, similar perceptions of the teacher labor market, and similar parental preferences by sector—meant that charter and TPS leaders shared and sought to promote similar interests: minimizing risk and appearing legitimate in a time of crisis. Charters did use their autonomy, but this was more relevant to how charters reopened, rather than when.

Other Governance and Institutional Factors

Although we did not find major differences in the timeline of reopening by sector within our case sites, we did identify important institutional differences between sites. The institutional context helped set the parameters within which actors negotiated their responses. Our case sites had different levels of centralized school governance and different degrees of coordination between school system leaders and other agencies and sectors. These differences are best illustrated by the reopening in New Orleans, where post-Katrina knowledge and capacity for crisis management supported its earlier school reopening timeline relative to our other cases.

Hurricane Katrina devastated New Orleans, and at the time it drew attention to the unpreparedness at the local and national levels (Committee on Homeland Security and Governmental Affairs, 2006). The city's historical experience with Katrina and other hurricanes before and since—and an institutionalized culture of disaster preparedness—greatly shaped its educational response to the pandemic. As one New Orleans principal put it: “Hey, if we lived through Katrina, we can pretty much live through anything.”

An important dimension of this institutional context was an openness to centralized decision-making. One school district leader explained how the city's prior experience with crisis management facilitated centralized decision-making that in turn supported its reopening timeline:

I think that our schools have autonomy that is unrivaled in any school district in the country. But I also think that our schools with that autonomy realize the limits of what it is that they can do well and what the limits are of like where innovation needs to happen. And innovation does not need to happen in the space of, “How many kids does the CDC think belong in a building?” And they’re sort of used to this like, “Okay. A hurricane's coming.”…And because of that, when you go into an emergency they understand that like, “This is the moment where I’m going to look to the district as the leader of this.” The same thing happened with COVID.

School system leaders used this central authority to establish common health and safety requirements for schools to reopen and as well as the city's first unified school calendar since its charterization. They viewed these steps as critical to facilitating in-person learning options—in the words of the district leader, to “get as many kids in person as possible that is safe and get as many people comfortable with that idea as possible.”

Another important element of New Orleans’ institutional context was coordinated support from the mayor. Other than in D.C., where the school system is governed through mayoral control, mayors did not play a prominent role in school reopening for our other case sites. One school system leader in New Orleans described the mayor's support for school reopenings like this:

For our city, our mayor has really been great. She's basically been, “My priority in all of this, is that as much in-person learning that can happen happens. Whatever you need, let me know. I, the mayor, don’t want to do anything to you that reduces your ability to have school…You need me to close bars so you can keep school open, we’re going to close bars so we can keep school open.”

The city did make a number of concrete decisions to mitigate COVID-19 case rates, including the cancellation of major events and conventions such as Mardi Gras and the Jazz Fest. School system leaders understood these actions were part of a commitment to prioritize school reopening, which helped with planning and implementing a phased-in reopening in the fall.

By contrast, we saw a lack of coordination in some cases. In Detroit, a fragmented institutional context meant limited collaboration among districts, civic leaders, and community organizations. For example, the TPS district mobilized philanthropic support to purchase and distribute technology to students, but there was no citywide effort to provide resources from the private sector to schools or to adjust health and safety rules for businesses. In addition, a coalition of community organizations coordinated among themselves to provide support to students, but largely had to do so without district collaboration. As one community organization leader explained, “We've tried and failed many, many, many times to build formal relationships.” D.C. also offers an example of limited collaboration: while the mayor's office coordinated leaders from the TPS district and charter school board with city officials near the beginning of the pandemic, collaboration with charter schools fizzled out entering into the 2020–21 school year. These cases highlight a clear distinction with New Orleans and its institutionalized culture of coordinated responses to crises.

Discussion

We began this study with two research questions. First, we asked how TPS and charter schools reopened in Denver, Detroit, New Orleans, Portland, and D.C., in 2020–21. In brief, we found that each city's TPS and charter school leaders chose to open through a phased process. We noted several important dimensions of reopening on which the cases varied, including the timeline for reintroducing in-person instruction, different kinds of in-person options, different health and safety measures, and choices offered to parents and (in some cases) teachers.

Our description of the nature and timing of reopening enhances the existing literature on reopening. The dominant political discourse often posed a binary question: will schools be open or closed? (Cohen, 2022). Prior research also tended to adopt this kind of framing, defining reopening or modality in limited ways or focusing on only one or two points in time. Our account of reopening these five districts reveals a much more dynamic process. What counted as “open” differed between cities, and in some cases among schools within the same city, based on health and safety measures in place, the groups of students given prioritized, and negotiated working conditions and instructional expectations for teachers. Further, once schools initially reopened, schools adjusted their instructional arrangements, including reclosing in some cases; and plenty of students remained in remote instruction even when in-person options were available. These meaningful differences are flattened by the “open or closed” framing and in binary or simple categorical quantitative data.

Second, we asked what factors shaped the school reopening decisions in each of these districts. We drew on micropolitical theory and used qualitative data, which allowed us to more closely examine the mechanisms through which relevant factors influenced reopening decisions. We investigated who the key actors were and what strategies they used to influence decisions. We also investigated whether and how these actors and their interactions were influenced by local political, institutional, cultural, and public health contexts. In this way, our micropolitical analysis significantly extends the findings in the prior literature and their implications.

Perhaps most importantly, our findings challenge the “politics or science” framing that has dominated both research and public discourse on reopening (Cohen, 2022). The polarized responses to the COVID-19 pandemic may be striking (e.g., Baccini & Brodeur, 2021; Kerr et al., 2021), but public health decisions have always been political (e.g., Oliver, 2006). We show that at every stage—especially during planning before fall 2020 but also throughout 2020–21—reopening decisions were both responsive to the local conditions of the pandemic and local political dynamics; in fact, the two were highly interrelated (Harris & Oliver, 2021).

By attending to the role and interests of local actors, their strategies, and the influence of context, our study builds significantly on the general finding that more heavily Democratic districts were slower to reopen schools. Illness and death from earlier waves of COVID-19, and national partisan responses, shaped the political climate for our case districts leading into the 2020–21 school year. When planning for reopening, school system leaders in our cases consulted with local health authorities and adhered to national guidelines because they wanted to legitimate their decisions in response to health concerns among their (heavily Democratic) constituents and make decisions in line with their own values and political outlooks. These plans, along with leaders’ attention to local COVID-19 case rates and the politics surrounding them, shaped further opening and closing decisions throughout the year.

Our findings also show the limitations of national partisanship explanations for school reopenings. Partisanship and national debates over reopening absolutely affected the strategies and actions of local stakeholders, reflecting the “end of exceptionalism” in education (Henig, 2013). Yet, the variations we found across our cases show that even during this national crisis, education policymaking retains much of its federalized character. Local actors and contexts continue to matter, and a micropolitical perspective is necessary to understand and explain educational decisions.

We also detail the influence of teachers’ unions and teacher labor market conditions more broadly. We found that unions in Detroit, Portland, and DC directly influenced the nature or timing of reopening, while in Denver their influence was less substantial. Further, school and district leaders—in both unionized TPS districts and (almost entirely nonunionized) charter schools—were broadly concerned about the stress of the pandemic on teachers and made reopening decisions with these concerns in mind. The correlations between union strength and reopening decisions in the existing quantitative studies may be picking up on some influences beyond the strength of unions or their actions in negotiation with their districts. For example, these measures might capture greater concerns for teacher shortages or vacancies, which could be caused by strong union presence or actions but could also plausibly be related to sociopolitical factors that made those districts amenable to strong unions in the first place. In addition to our findings, Kretchmar and Brewer (2022) found evidence that in their Republican case districts, decision-makers and the broader public expressed less concern about the threat of pandemic for teachers. This could indicate less sympathy for teachers overall, or it could be a specific product of views on COVID-19 and not indicative of their broader support for or animosity toward teachers. Indeed, Houston and Steinberg (2022) found that counties, where the public had greater support for teacher salary, increases pre-pandemic also had a larger share of their schools open; though, the authors note, they “do not identify a specific causal mechanism” underlying this association (p. 4). Thus, attitudes toward teachers and unions—including sensitivity to teacher well-being and concerns about shortages and vacancies—are likely a distinct part of the political context for reopening that merits further attention.

Another reason the micropolitical perspective is useful is in its attention to influence, and the extent to which actors strategically exercise or cede power in the decision-making process. In Denver, Detroit, and New Orleans, school and system leaders were at the center of reopening decisions because other actors who might have otherwise influenced their decisions instead gave them discretionary authority (e.g., state lawmakers, school boards). On the other hand, leaders of the executive branch played a decisive role both in keeping schools closed and reopening them in Portland (the governor of Oregon) and D.C. (the mayor). We also observed virtually no role for mayors in Detroit, Denver, and Portland. Although mayors in these cities do not have formal power over schools (as in D.C.), they might have sought to influence reopening by coordinating among community and business interests, as in New Orleans, or by using their bully pulpit to influence public opinion (Collins, 2021a).

Facing the politically difficult position of making school reopening decisions, school and system leaders exercised and ceded authority in strategic ways. Most consistently, they made strategic appeals to public health guidance to strengthen their positions as they decided to open schools (or keep them closed). Detroit and D.C. also had examples of charter leaders strategically ceding their authority somewhat, by deferring to the TPS district's decisions. We should also recognize the modes of communication that school and system leaders used to communicate with parents as a strategy. The use of surveys and town hall-style meetings had an organizational function, as district and school leaders sought information to use in their planning and allocation of resources; but these modes of communication were also political insofar as they shape the channels for parent voice.

Finally, unions varied in their strategies as well. In Detroit, the teachers’ union secured perhaps the most substantial demand of any of our case sites—the right of teachers to choose their modality—in part by authorizing a strike. In Denver, where the union was less successful in securing more authority over reopening (e.g., a COVID-19 case rate at which schools would automatically close), they did not. Given that unions may have had more influence over reopening decisions through their established bargaining and negotiation power rather than through new efforts to organize influence (Marianno et al., 2022), these tactical differences were likely consequential. At the same time, however, we had examples of charter schools that were not unionized but gave their teachers a significant voice in reopening decisions—a reminder that the context for decision-making is highly influential as well.

Our findings have some implications for how educational leaders might navigate future crises. We should be cautious about extrapolating from school reopening during COVID-19 to all crises, since the pandemic had particular features that not all crises will share (e.g., prolonged rather than punctuated, with an evolving understanding of the threat). If we face another similar crisis, however, superintendents and principals may find themselves similarly responsible for making high-stakes decisions. Indeed, some school leaders in the United States faced these kinds of pressures about a decade ago, during the H1N1 pandemic (Klaiman et al., 2011). Educational leaders should be aware of and prepared for this. One step might be to invest now in building cross-sector coalitions to facilitate coordinated multiagency responses (Henig et al., 2019); and participatory structures that would provide parents and teachers a direct channel of influence and help build a collaborative decision-making culture for schools and districts (Collins, 2021b; Gandin & Apple, 2002). These kinds of relationships and processes might enable more flexible, equitable, and dynamic approaches to decision-making in the face of complexity and uncertainty. Educational leaders also need to recognize and may need to act early in their role as public figures who can potentially influence decision-making (Collins, 2021a).

We note a few limitations of this study. Due to resource constraints and restrictions related to COVID-19, we were unable to obtain interview data from some actors that may have exerted an influence on reopening decisions, including private school leaders and leaders from other districts regionally. Similarly, we were only able to interview a limited number of parents, community-based organization leaders, and charter school leaders in each city. As such, those data may not capture the full range of interests and influence strategies for those populations. In addition, while we paid close attention to the actions and decisions of school system leaders, we did not collect data on personal characteristics, such as their own political ideologies or partisan affiliations, or professional self-identities. Thus, we could not examine whether leaders’ political and professional dispositions aligned with or differed from those of teachers and parents in their districts, nor could we examine how those dispositions influenced their decisions. Finally, since each of our case sites had strongly Democratic-leaning constituencies, Democratic mayors, and Democratic state governors, we were not able to explore differences between school systems with different political compositions (see Kretchmar & Brewer, 2022).

With these limitations in mind, there are several fruitful directions for further research. First, there is more to learn from how reopening was enacted in different contexts (e.g., Republican-led states and cities, rural districts); both detailed case studies and comparative research would be useful. Second, research should try to bridge the gap between school reopening (and factors that influenced it) and actual participation rates (of students and teachers) in in-person instruction. Such an analysis would help broaden our understanding of the issue itself. Further research should ask how the combination of factors identified in this study and others influenced the way students learned, including the learning modalities their schools offered and families’ decisions to participate in those. Our inquiry also shows that reopening is not simply a top-down matter controlled by school system leaders. Further research is especially needed into whether, how, and with whom parents engaged in individual and collective action seeking to shape reopening decisions, and what if any impact they had. Finally, further research on reopening and other decisions during the COVID-19 pandemic can adopt and build on the micropolitics perspective, to capture additional insights into educational decision-making at all levels during crises.

Author Biographies

Jeremy Singer is a postdoctoral research fellow with the Education Policy Innovation Collaborative (EPIC) at Michigan State University.

Julie Marsh is a professor of Education Policy at the Rossier School of Education and Sol Price School of Public Policy at the University of Southern California, Faculty Director of Policy Analysis for California Education (PACE) at USC, and Co-Director of the Rossier Center on Education Policy, Equity and Governance (CEPEG).

David Menefee-Libey is William A. Johnson Professor of Government and Professor of Politics at Pomona College.

Jacob Alonso is a PhD student in Urban Education Policy at the University of Southern California Rossier School of Education.

Dwuana Bradley is an assistant professor at the University of Southern California Rossier School of Education.

Hanora Tracy is a doctoral candidate in Sociology and part of the City, Culture, and Community program at Tulane University.

Appendix A

Parent and Guardian Interview Methodology

We interviewed parents and guardians who responded to an online, opt-in survey for parents or guardians of school-aged children across the five states and D.C. (For details on this survey, see Authors, 2021.) We selected interview participants from a list of survey respondents who, at the end of the online survey, indicated interest in participating in follow-up interviews. We indicated that interview participation was voluntary, confidential, and would result in a $25 Amazon gift card. Among those indicating interest, we further narrowed the sample to include only parents who lived in the five urban districts in our study.

Given our research focus on issues of choice and equity, and the disproportionate impact of the pandemic on historically marginalized families and students, we also believed it was crucial to oversample along with the lines of race and class. As such, we purposefully select interested parents from households earning less than $75,000 annually.

From this subset of parents, we then targeted parents representing variation in race and ethnicity (keyed to the prominent racial groups in each district), grade level of child (elementary and secondary), and school type (charter, traditional, private, and homeschool). These telephone interviews averaged 45 min in length and were audiotaped and transcribed.

Appendix B

School Modality Over Time, 2020–21 School Year

Modality data come from the COVID-19 School Data Hub (CSDH; covidschooldatahub.com). “Any in-person” indicates hybrid or in-person modalities offered. Blank areas indicate missing data. For Detroit, we used district-level data, which was more complete; Michigan collected modality data at the district level, and most Detroit charter schools operate in their own separate district. CSDH had no data for charter schools in Washington, D.C.; charter school data for quarters three and four are imputed from public records on charter school modality from the D.C. Public Charter School Board (dcpcsb.org/school-reopening-and-recovery). Public school data are left missing for New Orleans because only two schools were identified as non-charter in the CSDH.

Appendix C

Daily COVID-19 County Case Rate, March 2020 Through June 2021

COVID-19 data come from the New York Times COVID-19 data repository (github.com/nytimes/covid-19-data). Case rate (y-axis) is defined as COVID-19 cases per 100,000 residents. Vertical lines identify the first day of school for each case district.