___________________

¹ Student engagement is defined as the student’s active participation in academic and other school-related activities. See https://www.edglossary.org/student-engagement/ for a full definition.

learning. Many shifted to remote learning, with various levels of success. For young children, the social and educational benefits of early childhood education did not translate easily to virtual learning environments (Crawford et al., 2021). In K–12, the pivot to remote learning was complicated by the lack of time to prepare, limited availability of devices, and widely varying access to the internet (Leech et al., 2022). All public K–12 schools remained closed until the end of the 2019–2020 school year.

The capacity to shift to remote learning varied widely among schools and districts, depending in part on whether students and families had access to the internet and appropriate devices. There was no single approach, and much of the limited evidence available suggests that schools struggled, especially with student engagement. Absent preparation and support, of which there was little, teachers struggled as well (Barnum & Bryan, 2020). Unfamiliarity with Zoom and other platforms and adapting instructional strategies to virtual contexts proved challenging; as a result, the content being taught was not consistently engaging or rigorous (Lemay et al., 2021; Leech et al., 2022). As detailed below, these challenges were most evident for marginalized student populations.

There is some qualitative evidence that certain schools thrived, and those that did shared several attributes (Libette et al., 2020). First, they focused initially on student and family well-being, including through the distribution of meals and support in making connections to social service agencies. Second, they worked to ensure that students had the necessary technology to access virtual learning opportunities by working with internet providers and distributing computers and tablets. Third, they adapted their instructional practices to virtual environments and used approaches that better fit students’ and families’ needs. These adaptations included the creation of prerecorded lessons to watch at any time and the creation of small cohorts or pods to mimic the social structures that are commonplace in in-person schooling.

More generally, however, there were significant disruptions in providing quality instruction, whether in person or remote. For many students, virtual learning opportunities did not replace physical school. Rather, physical school closures meant no school—literally none at all, for days and even weeks (Kurtz, 2020; Lake & Pillow, 2022). And these challenges in operating schools had serious implications for students. The COVID-19 pandemic appears to have deepened the impact of disparities in access and opportunity facing many students of color in public schools, including technological and other barriers that make it harder to stay engaged in virtual classrooms.

The disruptions were especially difficult for some students. For many English learners, the abrupt shift to learning from home amid the challenges of the pandemic was an especially difficult struggle. LGBTQ+ students

faced particularly heightened risks for anxiety and stress and lost regular access to affirming student organizations and supportive peers, teachers, and school staff. These students also are at an increased risk of isolation and abuse from unsupportive or actively hostile family members. Most students experienced and have continued to experience some challenges to their mental health and well-being, and many have lost access to school-based services and supports, with early research showing disparities based on race, ethnicity, LGBTQ+ identity, and other factors (Jones et al., 2022; Sullivan et al., 2022).

Educators were also greatly affected, which gave rise to a variety of staffing challenges. Although those challenges existed prior to the pandemic, especially in low-income communities, they were made worse by a tight K–12 education labor market that has continued during the pandemic. Overall K–12 employment decreased by nine percent at the beginning of the pandemic (Bleiberg & Kraft, 2022). Data from the National Center on Education Statistics shed some additional light on what schools faced. More than 50 percent of schools reported that resignations were the reason for teaching staff vacancies; another 21 percent were due to teaching staff retirements (IES, 2022c). Public schools also reported that teacher absences have increased in comparison with previous school years. More than 75 percent of public schools reported that it is more difficult to get substitutes than it was before the pandemic. And almost 75 percent of public schools are frequently relying on administrators, nonteaching staff, and teachers on their free periods to cover classes (IES, 2022b).

Initial Effects on Early Childhood Education

When schools closed early in the pandemic, there were immediate sharp declines in early childhood program enrollments; these declines persisted well beyond the initial phases of the pandemic. These enrollment effects, which likely translated to missed learning and socialization opportunities, were not borne equally. The programs experiencing the highest enrollment losses were those serving low-income families, families of children of color, and families that did not speak English at home (National Institute for Early Education Research [NIEER], 2022). Furthermore, given that families shoulder more than half of the total costs of early care and education and that private costs increase with family income, the closure of early childhood education centers might have resulted in substantially reduced expenses among families with more resources (Bornfreund et al., 2020).

Shutting down child care and early learning programs also affected early childhood educators. Already poorly compensated, many reported experiencing financial as well as emotional stress (Crawford et al., 2021; Eadie et al., 2021). Early childhood programs gradually reopened after

initial closures, but variability in health and safety measures and reopening guidance and limited access to state and federal assistance, compared with the K–12 system, left many providers uncertain about when and how to reopen (Bassok et al., 2020; the National Academies of Sciences, Engineering, and Medicine [the National Academies], 2020b). Class size restrictions and the increased costs of mitigation measures, including the cost of cleaning supplies and personal protective equipment, put program providers under significant financial stress (Pattnaik & Jalongo, 2021; Schechter-Perkins et al., 2022).

The Challenges of Reopening Child Care Facilities and K–12 Schools

The effort to reopen elementary and secondary schools has had its own set of challenges. As the 2020–2021 school year began, some districts offered in-person learning; others offered only remote learning; and still others used hybrid approaches that combined the two, trying to balance the needs of safety and health with education. By January 2022, just 40 percent of K–12 students had access to any in-person instruction; by June 2022, more than 98 percent of public schools offered full-time in-person instructions (Institute of Education Sciences [IES], 2022).

The challenges involved in reopening schools for in-person learning ranged from the condition of educational facilities to the procedures for mitigating the ongoing public health challenges, as well as grappling with ongoing adjustments to organizing and delivering instruction.

Facilities

The condition of child care and education facilities was a concern well before the COVID-19 pandemic. A significant percentage of early childhood programs operate in private settings, including churches, neighborhood centers, and homes, many of which were in need of facility upgrades and improvements (Bipartisan Policy Center, 2019). For programs operating in traditional school settings, conditions mirrored the national infrastructure crisis; early childhood and K–12 school facilities were suffering from disinvestment (Petroski, 2016; Malik et al., 2018; Partelow et al., 2018). Deferred maintenance has been a growing problem nationally, with implications for both the deterioration of school buildings and their operating systems (Hunter, 2009). One response has been to engage in facilities assessments, in part to determine what is needed to safely reopen schools, but also to determine changes that might be needed if remote instruction or hybrid learning again becomes necessary.

___________________

² See https://oese.ed.gov/offices/education-stabilization-fund/elementary-secondary-school-emergency-relief-fund

would on average spend $55 per student for materials and consumables associated with these mitigation efforts (Rice et al., 2021). Added expenses for custodial activities and the additional costs for transportation accommodations raised these estimates to between $250 and $700 per pupil. The costs related to social distancing within classrooms were difficult to predict for reasons described below. However, many schools had to create physical barriers in classrooms, such as plexiglass shields at individual desks, when class size and classroom space limited their ability to create space between students. Understanding now that the virus is likely to continue in the population, some mitigation measures have remained in place in many schools, including disposable facemasks, gloves, hand sanitizers, thermometers, and signage.

School Management and Organization

One of the most important adjustments schools made as they reopened was reducing class sizes. Subject to wide variation by school level and type, the prepandemic average class size was approximately 19 students per teacher (IES, 2019). Larger class sizes mean more crowding and thus increased risks for contact and spread of the virus; bringing average class sizes down created greater opportunities for physical distancing.

In many instances, reducing class size required making changes to school schedules; one way this was done was by alternating the days of the week when specific grades came to school (CDC, 2022). This approach varied as schools navigated successive waves of the pandemic. A common approach had one group of students come to class on Monday and Thursday while a second group came on Tuesday and Friday. On days when students did not attend class, teachers gave them online assignments. Some schools used the middle of the week (e.g., Wednesdays) to allow for one-to-one tutoring and a variety of other interventions to address the unique challenges of virtual or hybrid learning experiences, such as problems with devices and connectivity and access to hard-copy materials, projects, and extracurricular activities (CDC, 2022b).

It is important to note that over the past two decades schools have become a primary source of meals for children. When the pandemic forced schools to close, many districts initially focused on meeting the critically needed meal distribution.³ Once open, many schools found it necessary to stagger meal periods throughout the day or have students eat at their desks,

___________________

³ In 2020, 22.6 million children participated daily in the National School Lunch Program: see https://www.ers.usda.gov/topics/food-nutrition-assistance/child-nutrition-programs/national-school-lunch-program

rather than in a usually crowded cafeteria, with appropriate protocols to keep the classroom clean.

Closing and reopening schools and child care facilities was a challenging task, requiring attention to a wide range of issues, from access to facilities to the organization and management of classrooms. Opening schools was further complicated by the onset of new variants and the ensuing controversies over masking requirements, student and teacher safety, and frequently changing public health guidance. There has been significant variation in how students experienced education during the pandemic, with implications for equity. This in turn has pushed schools to begin and continue to mount a variety of efforts to address what they view as missed opportunities to learn.

Instructional Modes

The American Institutes for Research (AIR) conducted two surveys to study how schools responded to COVID-19, the first in spring 2020⁴ (AIR, 2020). The survey found that, on average, students in the early grades were expected to devote 2–3 hours a day to instructional activities. Students in middle schools (grades 6–8) and high schools (grades 9–12) were expected to spend 3–4 hours each day. Since most states require 5–6 hours of instruction each day (IES, 2020), these time expectations fell short of state regulations. There were modest differences between high- and low-poverty districts with 55 percent of high-poverty districts requiring in-person instruction compared with 47 percent in low-poverty districts (AIR, 2022).

The AIR survey also found variations in the types of activities emphasized in the context of remote instruction. For some districts, remote instruction meant reviewing content that was covered earlier in the year. For others, students were exposed to new content. Again, on average, low-poverty districts reported spending less time reviewing previously taught material than their high-poverty counterparts. These differences in district wealth also translated into differences in how remote instruction was delivered and supported. Wealthier districts were more likely to provide printed materials to support remote lessons, more likely to have classes that were taught by the teacher of record, and more likely to have virtual lessons that were prerecorded and available asynchronously for students and parents: see Figure 4-1.

More research is needed to understand how school districts and schools approached instruction at the beginning of the pandemic. It is clear,

___________________

⁴ The sample sizes for this survey were 454 districts for grades K–5, 454 districts for grades 6–8, and 432 districts for grades 9–12.

however, that schools and districts have faced a variety of challenges as the pandemic has continued and that both the existing technical infrastructure and the socioeconomic status of the district figured prominently in how it has responded and continues to respond to COVID-19. It is likely that as districts have had more experience with the pandemic, that experience has translated into more effective approaches to remote and hybrid instruction.

By winter 2021, in-person instruction was more widespread in schools. The second AIR survey found that just over one-third of the districts in their sample were mostly in-person with another one-third reporting that they were partly remote and partly in-person (AIR, 2022). This survey revealed that as in-person instruction increased, so did district expectations about the time students were expected to spend on instructional activities. The data suggest that instructional time is back to and in some instances now exceeds what most states require. However, the effects of school closures were significant (Conto et al., 2021); the next section presents the evidence on the pandemic’s effect on learning outcomes and other measures of student success and well-being.

STUDENT OUTCOMES

The pandemic’s effects on schools’ ability to function, as well as its broader societal effects, affected a wide variety of student outcomes across all ages. These outcomes can be grouped into two main categories: school engagement and learning. School engagement includes such measures as enrollment, attendance, graduation, and matriculation to postsecondary education. Learning includes numeracy and literacy, typically measured by standardized test scores. Across all measures, students appear to be worse off than they would have been absent the pandemic, and such negative outcomes are generally more severe for the low-income and racially and ethnically minoritized communities that are the focus of this report.

School Engagement

Engagement in schooling, from preschool through college, has dropped according to a variety of measures, including enrollment, attendance, and graduation. The available evidence, discussed below, suggests that because of the pandemic, fewer students are enrolled in public or any formal schooling or preschool; daily attendance is lower than it was prior to the pandemic; and high school graduation and college enrollment rates are lower than would have otherwise been expected.

experiences, and found it difficult to attract and retain qualified staff given the prevailing low wages in that sector. One result of the decreased demand and the supply constraints is that many child care centers have permanently closed. From December 2019 to March 2021, nearly 9,000 child care centers closed in the 37 states for which data were available, representing a nine percent decrease in licensed centers (ChildCare Aware of America, 2022).

K–12 Enrollment

Nationwide enrollment in public K–12 schools dropped substantially in the first school year after the pandemic began. Overall enrollment during the 2020–2021 school year was lower by 1.3 million students than in the 2018–2019 school year, a decline of nearly 3 percent. Four-fifths of that decline was concentrated in elementary grades. Kindergarten enrollment was particularly hard hit, with 9 percent fewer students enrolling than in the year before the pandemic. Grades Pre-K through 8 saw an average of a 3 percent decline in enrollment. In contrast, grades 9–12 saw a 0.4 percent increase (IES, 2021).

Overall enrollment declines varied by race, with White and Black student enrollment dropping by 5 and 3 percent, respectively. Latino student enrollments rose slightly, less than one percent, perhaps because Latino individuals are the fastest-growing demographic student group. Importantly, the large kindergarten declines observed were even larger in fully remote school districts, which disproportionately enrolled low-income students (Korman et al., 2020; IES, 2021).

Data from individual states provide further insights into which students left public schools in the fall of 2020 and where they went. In Virginia, kindergarten enrollment drops were largest among students who were low-income, Black, or Latino. Low-income students’ enrollment dropped by 32 percent, compared with a drop of only seven percent for non-low-income students. In Massachusetts, public school enrollment fell by almost four percent, more than 10 times the largest annual change in enrollment the state had experienced over the previous decade. This decrease was a combination of a 4.5 percent enrollment drop in traditional public school districts, a 2.7 percent increase in charter schools, and a 21.5 percent increase in virtual school districts (from a low base; Bassok & Shapiro, 2021; Dee & Murphy, 2021).

Student-level data from Michigan further illuminates where students were enrolling. In the fall of 2020, Michigan public school enrollment dropped by three percent and kindergarten enrollment dropped by 10 percent. The declines in kindergarten enrollment were highest among low-income and Black families. However, the declines in other grades were highest among higher-income and White families, which may highlight the importance of students’ attachment to their current schools. Most students who left school, particularly in elementary grades, moved to home schooling. Interestingly, home schooling increased more where schools provided

in-person instruction, while private schooling increased more where instruction was remote. This pattern is consistent with some parents being concerned about the health risks of in-person pandemic schooling, and other parents being concerned about the instructional quality of remote schooling. Data from the U.S. Census Pulse Survey show that rates of home schooling rose dramatically among low-income and Black families, consistent with fears in those communities about the dangers of COVID-19 in schools (Musaddiq et al., 2022).

Data for fall 2021 from more than 600 districts in 23 states showed that very few districts had returned to prepandemic enrollment levels. Most districts had seen a second year of decline. Nationwide, public school enrollments in fall 2021 were nearly identical to those in fall 2020, still 2–3 percent lower than in fall 2019 (IES, 2022b). Most districts had seen a second year of decline. Large school systems saw particularly large second-year declines, with New York City losing 38,000 students in 2020 and another 13,000 in 2021, Los Angeles losing 17,000 in 2020 and another 9,000 in 2021, and Chicago losing 14,000 in 2020 and another 10,000 in 2021 (Kamenetz et al., 2021). Projections for the 2022–2023 school year suggest even further declines in some large school districts, including the projected loss of 30,000 in New York City; a continuing downward trajectory from prepandemic projections in Washington, DC; and a third year of enrollment declines in Seattle, after years of stable or growing enrollment. These trends suggest that public schools continue to struggle to convince parents that school buildings are as safe and that instructional quality is as high as they were prior to the pandemic (Bamberger, 2022; Stein, 2022; Velez, 2022).

The national picture of where the students who have left public schools have gone is not clear. Data on enrollment in alternatives, such as private schools and home schooling, have come only from a few individual states and rarely in a timely fashion. The implications of declines in public school enrollments depend in part on whether such missing students are receiving high-quality education elsewhere, which is not known from the currently available data.

Attendance

Multiple sources of evidence suggest that, among enrolled students, K–12 school attendance was substantially lower in the 2020–2021 school year than in prior years. Data from PowerSchool, a company that helps school districts track student progress, showed that between 2019–2020 and 2020–2021 attendance fell in about 75 percent of its 2,700 school districts, by an average of 1.5 percent. EveryDay Labs, a company that helps 2,000 schools track attendance, found that the number of students missing

more than half of the school year was seven times higher in 2020–2021 than in the prior school year. It also found that absenteeism was greater among low-income students and English-language learners. Data from individual states are sparse. Connecticut reported that between 2019–2020 and 2020–2021 attendance dropped by 2 percentage points on average but by 4 percentage points among “high-need” students. Chronic absenteeism was higher among low-income, Black, and Latino students. It was lower among districts offering predominantly in-person instruction (Korman et al., 2020).

In surveys, 18 percent of school districts reported that attendance was much lower in fall 2020 than in fall 2019, with average reported rates 2–3 percentage points lower than prepandemic averages. Schools in high-poverty districts and districts serving mostly students of color were also more likely to report that fall 2020 attendance was substantially lower than in the previous year. Reported absenteeism was also higher among districts that primarily offered remote instruction (Carminucci et al., 2021).

Chronic absenteeism, defined as missing at least 10 percent of the school year, worsened dramatically during the pandemic and shows no signs of abating. In the 2021–2022 school year, 72 percent of public schools nationwide reported chronic absenteeism rates higher than the previous year, with only 27 percent saying the situation had improved relative to the 2020–2021 school year (IES, 2022b): see Figure 4-3.

In the 2021–2022 school year, 40 percent of New York City students were chronically absent, compared with 26 percent prior to the pandemic. In Los Angeles, the chronic absenteeism rate rose from 19 percent prepandemic to 46 percent in the 2021–2022 school year. Similar patterns are evident in smaller school districts, including Akron, Ohio; Richmond, Virginia; and Jefferson County, Kentucky (Camera, 2022). Even more extreme examples of absenteeism have been reported: almost half of public school teachers reported having a student on their 2020–2021 enrollment roster who never showed up for class. Teachers were substantially more likely to report such students if they were teaching in schools serving high numbers of low-income, Black, or Latino students (GAO, 2022).

Though some of this absenteeism may be explained by exposure to COVID-19, the magnitude is sufficiently large to suggest other factors keeping students out of school, including mental health challenges or more general difficulties coping with disruptions to school and home life. Even at the height of the Omicron wave in January 2022, over three-fourths of schools had fewer than 10 percent of students required to stay home for COVID-related reasons (IES, 2022b). Such quarantines, which typically lasted for 5 or 10 days, are unlikely to explain the rise in chronic or even more extreme forms of absenteeism documented above.

were only somewhat concentrated among Black, Latino, and Native American students. Between 2019 and 2021, freshman enrollment across all sectors declined by 14.3 percent for Black students, 8.6 percent for Latino students, and 19.6 percent for Native American students, compared with a 12.0 percent and a 3.5 percent decline for White and Asian students, respectively. For public 4-year colleges, relative to White students, enrollment declines for Native American and Black students were twice as high, while for Latino students the decline was only one-sixth of that for White students. In the community college sector, declines for Latino, Black, Native American, and White students were all large and roughly similar in magnitude (17–27%; NSCRC, 2022).

___________________

⁵ Dynamic Indicators of Basic Early Literacy Skills; see https://dibels.uoregon.edu

benchmark was the largest single-year increase in the history of PALS (Markowitz et al., 2021). The increases in students at high risk for reading failure were largest among Black, Latino, and low-income students, as well as English-language learners (Herring et al., 2021).

K–12 Test Scores

Test scores from fall 2020 suggest that the effects of the pandemic on student learning began early. Nationwide data from more than four million students in grades 3–8 who took the MAP (Measures of Academic Progress) Growth⁶ assessments of the Northwest Evaluation Association (NWEA) showed mathematics test scores 5–10 percentiles (0.13–0.25 standard deviations) lower than would have been expected from prepandemic levels. Overall reading scores remained relatively stable, though declines were observed among Black and Latino students in upper elementary grades. However, the challenges of administering standardized tests in fall 2020, particularly among socioeconomically disadvantaged students, meant that missing data complicated measurements of such disparities.

The fall 2020 administration of Ohio’s 3rd-grade English language arts assessment found that achievement had declined by the equivalent of one-third of a year of learning (0.23 standard deviations) from 2019. The proportion of students scoring at levels appropriate for promotion to 4th grade dropped by 8 percentage points. Black students lost closer to the equivalent of 1.5 year of learning, while low-income students also lost more ground than their higher-income peers (Kuhfeld et al., 2020b; Lavertu, 2021).

The end of the 2020–2021 school year saw further evidence of missed learning. Data for 12 states on proficiency rates on standardized exams in spring 2021 showed large learning reductions in comparison with prior years: students were 14 percentage points less likely to score at least proficient in mathematics and 6 percentage points less likely to score proficient in English language arts. Declines in achievement were significantly larger in school districts with a larger proportion of low-income, Black, or Latino students (Halloran et al., 2021). Comparing achievement growth for 3rd–8th-grade students from fall 2019 to fall 2021 to a similar prepandemic two-year period shows that the average student learned substantially less than predicted, with Black and Hispanic students’ learning roughly 50 percent less than the learning of White students (Goldhaber et al., 2022b).

The 2021–2022 school year saw some rebound in achievement, but not for all students and not at a pace sufficient to offset the losses of the preceding 2 years. Nationwide data from NWEA suggest that students

___________________

⁶ See https://www.nwea.org/map-growth

in elementary grades appear to have made up about one-quarter of their missed learning over the 2021–2022 school year. This pattern is consistent with state test results, which showed partial rebounds among elementary students in Texas, Florida, Tennessee, and other states. However, NWEA data suggest that middle school students have shown little rebound, with students in 7th and 8th grade making little progress and even falling further behind in mathematics. The income gaps that widened during the pandemic have closed only a bit in reading (with high-poverty schools making faster progress than low-poverty ones) and have not closed in mathematics. Progress in returning students to prepandemic learning expectations has been mixed and arguably too slow to undo the missed learning in a timely fashion, particularly for older students (Barnum, 2022; Kuhfeld & Lewis, 2022).

Finally, the National Assessment of Educational Progress (NAEP) reading and mathematics assessments given to nationally representative samples of 4th- and 8th-grade students showed unprecedented declines in mathematics and significant dips in reading achievement between 2019 and 2022. These declines were broad-based—affecting students in every state and every region of the country: see Figures 4-5 and 4-6. The average 4th-grade mathematics score decreased to its lowest level since 2005, and the average 8th-grade mathematics score decreased to its lowest level since 2003. Reading scores declined, though not as much as mathematics scores. Results for the assessments reflect the performance of students attending public schools, private schools, Bureau of Indian Education schools, and U.S. Department of Defense schools (IES, 2022b).

The missed learning documented above was driven by a combination of the pandemic and policy responses to it. Students, educators, and families all struggled with an unprecedented shock to home and school life. School closures were one potential contributor to missed learning. Declines in mathematics and English language arts proficiency rates on state exams were substantially smaller in districts that offered more in-person learning, relative to neighboring districts, with English language arts learning particularly sensitive to instructional mode for Black and Latino students (Halloran et al., 2021). NWEA data suggest that schools that continued in-person instruction saw little or no widening of test score gaps by race and income, while districts that had remote instruction saw larger declines in higher-poverty schools than lower-poverty schools. Most of the widening racial gap in scores during this time was driven by Black and Latino students attending schools where in-person instruction diminished, not by widening gaps between students attending the same schools (Goldhaber et al., 2022b). The 2022 NAEP results suggest states that offered less in-person instruction in 2020–2021 had larger declines in 4th-grade mathematics and

modestly larger declines in 4th-grade reading and 8th-grade mathematics (NCES, 2022).

Prior studies of the relationships between academic achievement and both individual and national economic outcomes suggest substantial costs from students’ missed learning during the pandemic. One study early in the pandemic estimated that the spring 2020 school closures would likely lower students’ future lifetime income by about three percent, as well as lowering future U.S. gross domestic product (GDP) by about 1.5 percent (Hanushek & Woessmann, 2020). A more recent analysis based on the 2022 NAEP score release suggests that recent 8th graders will experience a 1.6 percent lower value of their lifetime earnings because of their decreased learning (Doty et al., 2022). This projection is roughly in line with the earnings losses experienced by cohorts of Argentinian students who lost nearly months of instructional time due to teacher strikes, a piece of evidence suggesting the causal effect of lost learning on students’ later economic outcomes (Jaume

& Willén, 2019). Specific predictions of lost earnings are based on numerous assumptions and thus come with substantial uncertainty, but it seems likely that the missed learning experienced during the pandemic will have lasting economic consequences for individuals and society unless effective and timely interventions are implemented (Hanushek & Woessmann, 2020; Viana Costa et al., 2021).

CONCLUSIONS

Conclusion 4-1: Across all measures of school engagement and learning outcomes, students appear to be worse off than they would have been absent the pandemic, and such negative outcomes are generally more acute for the low-income and racially and ethnically minoritized communities that are the focus of this report.

and secondary school education was provided through the Elementary and Secondary School Emergency Relief (ESSER) Fund, across these three relief packages (CARES, CRRSA, and ARP), directly to school districts based on the proportion of funding they receive through Title I of the federal Every Student Succeeds Act (ESSA; Jordan, 2022). In general, the purpose of the ESSER funds was to help schools and districts offset the learning that students missed during the pandemic (described as “learning loss” in the legislation), provide supplemental learning opportunities, and provide emergency support to schools most affected by the pandemic. Of note, the funding to states in each of these rounds of emergency funds, while generally for governors to spend on education at their discretion, has been used in part to provide funding to private K–12 schools in their states.

The $13 billion of ESSER funding in the CARES Act has already been spent, the $54 billion in the CRRSA ($54 billion) has to be spent by September 2023, and the $122 in the ARP must be obligated by September 2024 (Jordan, 2022). Although this total of nearly $190 billion may seem sufficient to meet the goals of ESSER, Shores and Steinberg (2022) used various models for learning loss assessment and estimated that it is a fraction of what would actually be needed to make up for the learning that was missed during the pandemic: they estimated the amount needed at $325 to $930 billion. Others have estimated that the ESSER funds received by high-poverty schools, including those that had remote learning for a significant proportion of time, may be sufficient for pandemic-related achievement losses, but schools will need ESSER funds for a variety of needs, with academic achievement loss being just one of them (Goldhaber et al., 2022b; Gordon & Reber, 2022). The ARP requires that schools spend at least 20 percent of allocated funds on academic achievement loss; a recent analysis of school districts’ plans found that, on average, school districts plan to spend 25 percent of these funds on academic achievement loss, with the rest planned for facilities (such as upgrading ventilation systems), technology, staffing, and mental and physical health (DiMarco & Jordan, 2022).

Thus, as high-poverty schools (especially those that continued remote learning during the pandemic) plan their spending of the last of the ESSER pandemic education funding, they will have to make difficult decisions between funding the need to address missed learning, fixing aging and financially neglected school buildings and facilities, bolstering school staffing, and implementing programs and support for students’ social and emotional development and mental and physical health needs.

Evidence-Based Interventions

Strategies for supporting learning beyond classroom instruction and addressing academic challenges of marginalized students are not new. For

decades, school districts and local educational agencies have implemented a variety of approaches, including tutoring, summer learning, extended day education, and mentoring (McCombs et al., 2014, 2020; Neild et al., 2019; Robinson et al., 2021). Many of these programs have been rigorously evaluated, providing useful evidence to inform strategies to respond to pandemic-related missed learning.

Evidence-based interventions are practices or programs that have documented empirical evidence of effectiveness based on research and evaluation. The ESSA requires that evidence be classified into strong (evidence based on randomized controlled trials), moderate (evidence based on comparison between a group receiving the intervention and a control group without random assignment), and promising (evidence based on correlations between the program and outcomes without a strictly defined control group). The availability of strong and moderate evidence about numerous interventions around the country provides an opportunity to inform both the selection of interventions and implementation details. Given the multiple alternatives to use the additional resources provided by ESSER funding, cost-effectiveness of one intervention relative to others is a critical aspect to consider (Neild et al., 2019).

Among the portfolio of interventions available to schools, strong evidence suggests that tutoring has positive effects on school attendance, academic achievement, and social and emotional outcomes. A recent meta-analysis of randomized controlled trials suggests a positive effect of tutoring on learning amounting to 0.37 standard deviations on average (Nickow et al., 2020). This effect is very large in comparison with a wide range of educational interventions, including reduced class size, vacation academies, and summer schools (Kraft & Falken, 2021). Furthermore, tutoring appears to be especially effective to improve learning among students from low socioeconomic backgrounds (Dietrichson et al., 2017), providing an opportunity to reduce learning gaps widened by the pandemic.

Strong research evidence also provides information about which implementation features of tutoring programs might be particularly effective to compensate for extended missed learning (Heinrich et al., 2014). These features include high dosage (tutoring occurs at least three times a week), close alignment between tutoring and the curriculum, small group settings with tutor–student ratios of not more than one to five,⁷ and tutoring that occurs during the regular school day rather than after school (Fryer & Howard-Noveck, 2020). An additional feature that has not been tested in the context of randomized evaluations but that might be particularly relevant in a

___________________

⁷ Although one-on-one tutoring might generate greater learning gains, it is not sufficiently cost-effective given the need to scale up programs.

pandemic context is the need for tutoring programs to target specific areas of loss that promote further learning, such as numeracy and literacy.

Evidence supporting other interventions—including summer learning programs, extended school day interventions, and mentoring—is somewhat mixed, with some studies finding strong positive effects and others finding limited effects (Kraft, 2015; Dorn et al., 2020; Beach et al., 2021; Cruz et al., 2022; Lynch et al., 2022; Weiss et al., 2022). In all these interventions, two factors appear to strongly shape program effectiveness: the dosage of the intervention and its consistency over time. Interventions that are infrequent and sporadic tend to show limited or null effects, while consistent and high-dosage interventions tend to show substantial learning improvements (Kraft & Falken, 2021). These findings suggest that a critical criterion for deciding which intervention to implement is to make sure that resources are available to support adequate dosage over a sufficient period of time.

Research suggests that learning gains from summer learning programs are greater when programs last at least 5 hours per week and 5 weeks during the summer and when the programs are offered over multiple summers. This is particularly the case for programs serving low-income and racially and ethnically minoritized students, who are more likely than others to have experienced multiple stressors and sources of disruption during the pandemic. In the case of tutoring and mentoring, a critical determinant of effectiveness is the stability and consistency of the relationship between students and tutors. A stable relationship supports trust and promotes a deeper understanding of specific students’ needs by the tutors. Consistent relationships are a particularly important feature as a source of stability that could counteract the disruption, and in some cases trauma, experienced by students during the pandemic.

Flexible Portfolio of Diverse Interventions

Given the sharp reduction in learning during the pandemic that disproportionately affected marginalized groups, an approach that integrates several of these approaches simultaneously in order to provide comprehensive support to students could be particularly useful. Deploying different approaches together could be useful not only to compensate for specific areas of missed learning, but also to address consequences of the pandemic on students’ mental health, social and emotional development, loss of friendship networks, and school disengagement. These interventions could also provide a venue to detect early warning signs of disengagement or learning challenges. An additional consideration when designing a portfolio of interventions is the positive externalities these programs could have for families and caregivers. For example, extended day and summer learning programs have been shown to provide safe and supervised places for children during

the time when parents and caregivers are at work, enhancing labor force participation and potentially addressing some of the negative labor market consequences of the pandemic on some population groups (Beach et al., 2021; Kraft & Falken, 2021).

Support for Student Attendance

Student attendance is a critical factor for such programs as summer learning and extended school days. Research indicates that students who attend summer or extended-day programs regularly achieve substantial learning gains. In contrast, those who attend sporadically experience minimal or no gains. Several barriers hamper students’ continuous attendance, including lack of transportation, need to provide child care for younger siblings, need for income from summer jobs, and the sense that programs are just “more school.” For example, a randomized evaluation of a voluntary full-day 2-year summer learning program in five urban school districts found that each summer approximately 20 percent of students signed up but did not attend a single day the first summer, as many as 48 percent did not attend a single day the second summer, and a large number attended only sporadically (McCombs et al., 2020).

These findings highlight the importance of strategies to promote continuous attendance. Promising strategies include discussing benefits of the program during the recruitment period and providing meals, transportation, and incentives, such as gift cards. In order to combat a perception that summer programs are just “more school,” a promising alternative is to establish partnerships with community institutions, such as local YMCAs, recreation centers, and Boys and Girls Clubs, in order to provide sports and recreational activities alongside academic enrichment. These strategies are particularly important to make programs sufficiently attractive for children and families in underserved communities.

Address Challenges of Scaling Up Programs

Many of the programs described above were created for specific groups of students, but the vast educational disruption of the pandemic means that interventions would need to be scaled up significantly. Scaling up these interventions poses challenges in terms of infrastructure, staffing, and logistics. Provision of services that are intermittent or of poor quality might worsen student outcomes by accelerating disengagement and stress and damaging trust and relationships between educational providers and students and families.

In this context, attempts to scale up programs might be more successful if they rely on incremental expansion and if they are part of a

comprehensive effort to integrate the specific program in established structures of the public school system. In terms of program implementation, scaling up highlights the need to ensure that program delivery actually occurs at sufficiently high doses and with sufficient consistency over time. This objective is as important as the specific content of a program.

Scaling challenges are exacerbated by the staffing shortages across public schools resulting from the pandemic, which is particularly severe in some regions of the country (IES, 2022b). To address these challenges, schools and school districts might need to consider flexible and creative strategies to staff programs. Given that the skills required for tutoring or mentoring are different from skills needed for classroom instruction, staff other than teachers might provide a desirable alternative. In the case of tutoring, for example, paraprofessional tutors might be as effective as teachers and might be a much more cost-effective strategy.

Another alternative is to enroll volunteer tutors—including parents, community members, and older students—provided that they receive substantial training, oversight, and coaching. Creative strategies in which higher-grade students tutor younger students—such as having high school students tutor elementary students as an elective course or having college students tutor middle school students as a federal work-study job—have proven successful (Fong, 2021; Kim et al., 2021; Nickow et al., 2021) and could be used to scale up programs. As has been shown by experience, a central concern with a volunteer workforce is not the ability of tutors, mentors, or coaches to provide content to students, but rather their commitment to providing services continuously and consistently enough for high-dose teaching. Using volunteer educators presents challenges in terms of training, supervising, implementing background checks, and coaching, but the federal resources provided by the ESSER fund could provide the opportunity to implement volunteer training preparation to address these challenges. A potential positive long-term consequence of putting these systems in place could be the ability to build processes and infrastructure necessary to incorporate volunteers as tutors, mentors, or coaches as a long-term strategy.

Integrating Universal and Targeted Interventions

Crafting a strategy to support students’ learning requires making hard choices that consider the cost-effectiveness of different interventions and tradeoffs between individualizing services and the objective of serving large numbers of students. These choices suggest the need to combine universal interventions with targeted interventions. For example, given the proven effectiveness of high-dosage tutoring, it is plausible to implement this approach as a school-wide activity available to all students, woven into the school day, rather than a targeted intervention that occurs after the school

day. Universal provision of interventions such as tutoring would help to reduce its stigma and could contribute to student engagement and sense of belonging. At the same time, other strategies, such as mentoring, could be implemented as targeted interventions, primarily available to children with specific needs or early warning signs of dropout, attendance problems, or learning challenges (O’Cummings & Therriault, 2015).

Compile and Share Experiences with Strategies for Promoting Learning

Since the pandemic began, schools and school districts have developed and implemented diverse initiatives for promoting learning and mitigating learning losses (Allensworth & Schwartz, 2020). These initiatives provide valuable information and guidance about best practices to adopt and mistakes to avoid. A repository of these experiences could be a useful tool for helping support states and schools as they implement their own programs; such a repository could provide toolkits, training, evaluation, and quality improvement tools.

Schools would benefit from a systematized centralized repository of guidance and toolkits as they implement and update their own approached based on their experiences. Such a repository would provide training and coaching, evaluation, and quality improvement tools, among other resources. It would have to be free, user friendly, and updated regularly. Several initiatives already in place provide useful blueprints.⁸

Reengaging Students and Families and Restoring Enrollment

As detailed above, public school enrollments dropped nationwide during the pandemic, with the sharpest declines in the earliest grades. There are a variety of potential reasons for the decline, including family decisions to home-school, students needing to work because of family financial circumstances, and students being homeless. Students being behind in credits toward graduation is another significant reason. Given that children experienced the pandemic in a many different ways, reengaging students will require a combination of efforts that address the many reasons for inconsistent attendance or disenrollment.

Collect Better Information About Enrollment and Engagement

The first step in getting children reenrolled is gathering good data on school enrollments. Having such data makes outreach and engagement

___________________

⁸ See, for example, https://studentsupportaccelerator.com/, https://www.sagaeducation.org/saga coach, and https://proventutoring.org

efforts more effective, and makes it possible to effectively monitor attendance. The disparate effects of the pandemic demonstrate the importance of collecting and analyzing data at multiple junctures in order to detect patterns that might affect only certain groups of students. By extension, expanding metrics gives school officials more information for designing appropriate interventions.

To the extent that returning to school includes hybrid approaches, monitoring attendance in remote settings will also assist with effective student engagement. There is growing evidence that virtual learning worked well for some students but not for others (Povich, 2020; Anderson, 2022). Better data on engagement in remote settings can help school officials to make decisions about which students should have priority for in-person learning opportunities.

Monitor Student Experiences and Outcomes

In addition to the usual challenges associated with children’s school attendance, the pandemic has brought about a number of additional stressors. There are students who are still struggling with the grief of the loss of a parent or close relative. For other children, the pandemic triggered stress and social anxiety. Older students who worked during the pandemic may have found agency in that experience, relative to how they experience school. Still other students may have inconsistent attendance due to feeling unsuccessful in the school environment; rates of course failures increased significantly during the pandemic (Belsha, 2022).

At a time when students have been affected by the pandemic in so many different ways, systems that detect and analyze the student experience create opportunities to intervene at both individual and group levels. Early warning and on-track systems are useful because they can identify students whose attendance patterns are problematic or who might be struggling (Heppen & Therriault, 2008; O’Cummings & Therriault, 2015). Predictive indicators of these sort of early warning systems provide help in identifying students as problems arise. These data can be used to detect patterns and trends so that school systems can address them proactively.

These types of systems have been in use for some time, primarily as a dropout prevention strategy. In 2014–2015, over one-half of all high schools across the country had early warning systems capable of identifying students at risk of educational failure (U.S. Department of Education, 2016). Adopting them for use for all students as part of a pandemic recovery strategy appears to be feasible and to have merit.

Other strategies for mitigating inconsistent attendance and chronic absenteeism include the use of periodic student (and family) surveys to obtain data on how students are experiencing school. Survey instruments

are available that elevate students’ voices on issues of school climate and their sense of belonging in schools.⁹ States and districts that have access to these types of data note that they have a much better understanding of how students are experiencing school and are nimbler in making adjustments that respond to student needs (Gross & Everett, 2021).

Strengthen Relationships Between Schools, Children, and Families

In addition to collecting good data on the students’ experiences, providing the right supports to students when they need them is critical. Bringing students back to school is, some argue, a relational challenge as much or more than a technical challenge (Student Education Equity Development Survey, 2021). Bringing students back to school and addressing the issues that led to their absence requires strong personal connections among students, teachers, and staff (CDC, 2022).

One approach to establishing personal connections between students and adults is reflected in Nashville’s Navigator Initiative, in which teachers and other school staff meet regularly with students to get to know them better and to learn what kind of support they need (Jacobson, 2020; Metro Nashville Public Schools, 2021). Navigators discuss topics that range from academics and enrichment activities to basic needs, such as food and housing. As with other interventions, training and tools (e.g., scripts for conversations and lists of available resources) for the adults contribute to success.

Strengthening the Educator Workforce

For years, recruiting and retaining qualified staff has been an issue in both early childhood and K–12 education (Totenhagen et al., 2016). In the context of early childhood education, COVID-19 appears to have had profound effects on educators as their schedules and working conditions were drastically and negatively changed. These stressors translated into large increases in teacher turnover (Weiland et al., 2021). Failing to address these staffing challenges is likely to compromise the quality of early childhood education for the long term.

In K–12 education, staffing shortages predate the pandemic for a variety of professions, including teachers, guidance counselors, and school nurses. Teacher turnover is a substantial problem, with more than 300,000 teachers leaving each year, most before retirement age (Sutcher et al., 2016; American Federation of Teachers, 2022; National Conference of State

___________________

⁹ See, for example, https://youthtruthsurvey.org/, https://www.panoramaed.com/panorama-student-survey, and https://www.oregon.gov/ode/educator-resources/assessment

Legislatures, 2022). Factors that contribute to high turnover include stress, lack of support, low pay, and heavy workloads (Sutcher et al., 2016). The pandemic exacerbated these issues and added additional stresses; teachers are leaving the profession long before they had planned to, and the pipeline of new teachers has shrunk (Camera, 2022; García et al., 2022; IES, 2022b; Walker, 2022).

Given the prepandemic conditions, along with the additional stresses of the pandemic, identifying viable strategies for strengthening the educator workforce will be key to addressing other policy interventions in education.

Improve Recruiting Efforts

To strengthen the educator workforce, it will be critical to improve recruitment of new teachers, counselors, and other education personnel. One promising recruitment approach is targeting paraprofessionals and teachers’ aides. Already in the schools, these individuals often live near the schools where they work, know the students and families, and may already be involved in efforts to foster engagement. The challenge is to design initiatives that support paraprofessionals in completing the requirements to become teachers or other licensed personnel. Some examples in the field highlight the financial and professional supports needed (e.g., coaching and mentoring; Solis, 2004; Educators for Excellence, 2017; Maready et al., 2021).

Another approach for recruiting is to focus on and develop strategies that address the needs and interests of individuals with potential interest in teaching as a career option. Programs such as Teach for America have demonstrated a track record of attracting noneducation majors to consider teaching by emphasizing the intrinsic reward of teaching and demonstrating the opportunities to promote equity and social justice. Other strategies focus on the practical barriers to entering teaching, by offering, for example, student loan forgiveness for time on the job, rent or mortgage assistance for high-cost markets, supports for credentialing and advanced degrees, and the articulation of clear and compelling pathways for advancement and professional growth (Podolsky et al., 2016). In addition to these specific strategies, a critical factor shaping recruitment and retention of the educational workforce is the support of their economic, psychological, and professional well-being. As highlighted by a recent report, Addressing the Impact of COVID-19 on the Early Care and Education Sector (the National Academies, 2022), this includes investments in educators’ wages and benefits, provision of professional recognition bonuses, mental health support, and an improvement of overall working conditions.

___________________

¹⁰ See https://nces.ed.gov/pubs2020/2020103/index.asp

¹¹ See https://nces.ed.gov/programs/coe/indicator/cge/racial-ethnic-enrollment

___________________

¹² See https://www.whitehouse.gov/covidplan/

children to school and among those most supportive of mask mandates and mask-wearing among their own children (Mostafavi, 2020; Gewertz, 2022). Such families have also been disproportionately represented among those who found it hard to access or afford high-quality masks and rapid testing, in part because their children attended school districts that lacked the funding to provide those resources (Jimenez et al., 2021). Setting aside funding for such mitigation measures that could be quickly activated in the face of future disease surges would increase both the actual and perceived safety of school buildings.

The benefits of both ventilation and detection and mitigation investments will be threefold. First, they will allow schools to remain open and functioning even during potential future surges of COVID-19 or other infectious diseases. Second, they will reengage at least some families who disenrolled their children from schools because of health fears. Third, by making the educator workplace safer, they will help schools better retain and recruit the educators needed in the nation’s schools.

REFERENCES

This page intentionally left blank.