Planning for School Reopening and Recovery After COVID-19 - An Evidence Kit for Policymakers
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Planning for School Reopening and
Recovery After COVID-19
An Evidence Kit for Policymakers
Shelby Carvalho
Jack Rossiter
Noam Angrist
Susannah Hares
Rachel SilvermanCenter for Global Development. 2020. Creative Commons Attribution-NonCommercial 4.0 Center for Global Development 2055 L Street, NW Fifth Floor Washington, DC 20036 www.cgdev.org
Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii 1 Engaging Communities in School Reopening Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 Targeting Resources Where They Are Most Needed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 Getting Children Back to School . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4 Making School Environments Safe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 5 Recovering Learning Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Technical Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Many thanks to Radhika Bhula and John Floretta at J-PAL for contributing to the briefs, and to Laura Moscoviz and Ana Minardi for research assistance. The Center for Global Development is grateful for contributions from the Bill & Melinda Gates Foundation, the Elma Relief Foundation, the UBS Optimus Foundation, and the Vitol Foundation in support of this work.
Introduction
Most governments around the world have temporarily The briefs are accompanied by CGD’s COVID-19 edu-
closed schools in an attempt to contain the spread of cation policy tracker, which provides up-to-date
the COVID-19 pandemic. Many have launched distance information on each country’s policy response and
learning programs and are beginning to plan for recov- reopening plans, as they emerge.
ery, a phase that involves much more than reopening
The World Bank, the World Health Organization,
the gates and readmitting students.
UNESCO, UNICEF, Education International, the Inter-
During the response, there may be opportunities for agency Network for Education in Emergencies, and the
governments to improve practices and “build back World Food Programme have released comprehensive
better,” but decisions are being made under extreme guidance on education policy responses to COVID-19
uncertainty. In this context, education policymak- and planning frameworks for school reopening. This
ing is particularly difficult but will be strongest when evidence kit complements those resources.
it is: (1) informed by existing rigorous evidence;
In areas where the rigorous evidence is thin, we high-
(2) approached with a long-term perspective that prior-
light this and emphasize the importance of generat-
itizes flexibility, communication, and trust; and (3) able
ing knowledge during this pandemic to inform policy
to be adapted based on new data and information
adaptations now and in future crises. Each brief is
gained from community engagement and monitoring.
complemented by an in-depth technical appendix
The short and accessible briefs collected here draw with details on the evidence supporting key findings
on rigorous evidence relevant to the COVID-19 emer- and recommendations.
gency to formulate recommendations for policymak-
There is still a huge amount to learn about the impacts of
ers on five critical dimensions of school reopening and
each country’s education policy response to COVID-19;
recovery:
it is too early to speculate on the effectiveness of different
1. Engaging communities in reopening plans choices, or their potential in different contexts. We will
be keeping our eye on the evidence as it emerges and,
2. Targeting resources to where they are most
in the months ahead, we will release additional analyses
needed
of the latest evidence on promising practices and pre-
3. Getting children back to school paring for future emergencies. In the meantime, CGD’s
COVID-19 education policy tracker provides up-to-date
4. Making school environments safe information on each country’s policy choices.
5. Recovering learning loss and building
back better
ii1. Engaging Communities in
School Reopening Plans
Successful reopening and recovery rely on public trust
in the government. Trust can be built and maintained
Recommendations
by engaging communities in planning for reopen-
ing and by clearly communicating through credible To productively engage communities and
channels. build the trust needed to design and imple-
ment effective reopening plans, policymakers
Policymakers will want access to evidence that sup- should:
ports their planning and decision making, and to draw
on relevant experience from elsewhere. This brief • Prioritize community engagement early
summarizes available rigorous evidence regarding the to build trust, shape perceptions of risk,
engagement of communities in planning for school and improve responses to government
reopening and identifies two recommended actions policy.
for policymakers based on that body of evidence.
• Share clear, credible, and consistent
messaging through multiple channels
Summary of Evidence to reach all groups and to match local
resources and norms.
Trust, risk, and community participation
A robust body of evidence suggests that establishing
public trust, managing fear and perceptions of risk,
and leveraging community engagement are key to engaging teachers and teachers’ unions, school-
effective crisis response. Governments need to decide based management committees and other par-
how to engage community members in reopening ent organizations throughout can increase the
plans and implementation; what, when, and how often relevance of plans and compliance with govern-
information should be shared with affected commu- ment interventions as schools reopen.
nities; and with whom and through which channels to
communicate. Communication strategies
There is limited evidence regarding the effectiveness
The evidence on building trust with communities of specific communication strategies to support emer-
affected by emergencies shows that: gency response. What is known strongly suggests that
communication content should be clear and credible,
• Leveraging community involvement to build
and be disseminated in a manner aligned with local
trust will play a vital role in facilitating the
norms. The best communication channels for reach-
reopening process and in shaping public per-
ing households will vary considerably across contexts
ceptions of the state over the long term.
but, wherever possible, multiple channels should be
• Effectively responding to health crises requires used to reach all groups. The “Signal Code” outlined
localized efforts to work with communities in the humanitarian and disaster literature argues
affected by the crisis. For the education sector, that not only does the public benefit from consistent
1information about the crisis and reopening plans, it Many governments have already strengthened
has the right to it. radio, television, and mobile communications
during the crisis; these same channels could be
The evidence on effective communication strategies
used to deliver reopening messages. For exam-
following an emergency indicates that:
ple, SMS allows for direct communication to
• Moving quickly to establish regular communi- community members at speed, and mass mes-
cation channels and messages focused on facts saging might provide a low-cost method for pro-
and key messages can help manage fear during viding information, encouraging compliance,
response as well as facilitate school reopening. and monitoring outcomes of school reopening.
Several ministries of education are also using
• Information is best shared using sources judged social media to communicate with parents about
credible by communities. In the wake of COVID- COVID-19, but this approach will not be suitable
19, we have seen numerous ministries of edu- everywhere. Word-of-mouth via parent groups
cation combat misinformation campaigns. and locally nominated individuals remains an
Communities are much more likely to fol- important communication channel, particularly
low guidance from sources that are nationally where literacy and connectivity rates are low.
well known and which they consider credible,
although it is important to be aware that source • Research suggests that combining multiple
credibility varies among groups and even within methods of communication could increase
countries. uptake and influence public behaviour as
schools reopen. For example, television and
• Communication channels should be selected SMS messaging can complement one another
based on their potential reach and local norms. to strengthen a single strategy. Broadcasting
Several methods have been proposed as chan- the same information in multiple formats can
nels for communicating response procedures make messaging more coherent. But be aware:
and reopening plans. Household surveys and information that is inaccurate or conflicts across
local norms about information acquisition can modalities may lead to confusion and noncom-
be used to determine each method’s potential. pliance with reopening plans.
2 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for Policymakers2. Targeting Resources Where
They Are Most Needed
A strong and equitable reopening and recovery
requires the careful use of resources, which relies on
Recommendations
broad coordination, effective targeting, and continual
use of data to adjust and improve approaches. To help ensure that resources are targeted in a
manner that supports a strong and equitable
Policymakers will want access to evidence support- reopening of schools, policymakers should:
ing their planning and decision making, and to draw
on relevant experiences from elsewhere. This brief • Coordinate actions and resources.
summarizes the rigorous evidence available regarding Under these exceptional circumstances,
the targeting of resources to where the need is great- managing competing priorities requires
est and identifies three recommended actions for coordination and exchange across orga-
policymakers. nizations, which can provide incen-
tives that strengthen implementation of
reopening plans and improve efficiency.
Summary of Evidence
• Use existing administrative and survey
Coordinating actions and resources
data to identify risk factors and guide the
Coordination and resources are critical components
design of social transfer mechanisms.
of emergency response, but evidence is thin in these
Community inclusion and data privacy
areas. If systems are to “build back better” as schools
should be prioritized at all stages.
reopen, then resources available for education will
need to be maintained or, in some cases, increased. • Gather high-frequency data early and
During and after the crisis, resources will be pulled in continuously throughout the reopen-
many directions, both within and outside the sector. ing process to support implementation,
Under such exceptional circumstances, available evi- adaptation, and learning.
dence shows that the management of competing prior-
ities—and targeted allocation of resources—will require
strategic coordination across sectors and agencies.
(e.g., national, regional, local, and school lead-
The evidence on coordinating actions and resources ers). Evidence from the grey literature suggests
during an emergency indicates that: that focal points should be designated for each
level of response.
• Strategic coordination and clear communica-
tion between sectors and agencies can improve • Collaboration increases exchange and can cre-
the implementation of school reopening ate incentives that strengthen reopening pro-
plans. There should be clear channels of com- cesses. Evidence from past disaster recovery
munication and responsibilities across line min- efforts demonstrates that organizations bene-
istries as well as among levels of government fit by pooling knowledge and data, sharing best
3practices, capitalising on economies of scale, • When targeting support to particular groups,
and expanding reach while avoiding dupli- evidence from social transfers strongly suggests
cating efforts. In the many instances where that careful design and implementation of the
collaboration across ministries and partners mechanism is the key determinant of effec-
remains weak, collaboration around the reopen- tiveness. Any targeting approach will be imper-
ing of schools creates an opportunity for its fect, regardless of the inclusion criteria (e.g.,
strengthening. geographic or means-based). What matters is
the careful use of data from multiple sources,
• Rapid response guidelines prioritize the use of
such as to enable targeting at multiple levels
existing educational resources and infrastruc-
(e.g., regional, provincial, district, and school
ture in the short term over the building of new
community).
systems. Innovative solutions building on exist-
ing systems, including acceleration programs, • Failing to include those on the periphery of soci-
can support rapid school reopening within ety in response plans could risk their further
existing resource and infrastructure constraints. marginalization as schools reopen. Inclusive
community engagement is critical to disaster
Targeting and inclusion response. COVID-19 will not affect all people
School reopening strategies should use existing data to or communities equally. Several countries are
target communities with the greatest need. Education already using data-led strategies to identify out-
administrative data can be combined with informa- of-school children. Policy guidance related to
tion from national household surveys and other read- postdisaster education planning also highlights
ily available sources. These data should be used to plan how important it is to ensure that communica-
reopening activities based on underlying risk factors tion of risk-based information targets women at
and knowledge of local resources and norms. Evidence all stages.
from past crises highlights the need to ensure data pri-
vacy and protection at every stage. • Data privacy and protection must remain pri-
orities during crisis and response efforts. Evi-
The evidence on strategies that use administrative and dence from past crises suggests that the urgency
survey data to enhance targeting and inclusion shows of data collection in an emergency can lead to
that: poor practices. Effectively targeting services to
• School reopening plans should account for the most vulnerable children and households
underlying risk factors, including student may require collecting sensitive information,
attendance rates, percentage of economically which will heighten the need to prioritize pri-
disadvantaged students, student-teacher ratio, vacy and protection throughout the process.
average years of teacher experience, and the
Monitoring and adapting
presence of vulnerable groups—including girls.
Response monitoring used during the crisis, including
Indicators for several of these risk factors exist in
phone surveys, can also be used to monitor reopen-
annual school censuses and household surveys,
ing. Evidence from past crises supports community
which could be used immediately.
involvement in the monitoring and evaluation pro-
cess. There is limited evidence specifically related to
4 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for Policymakersthe monitoring of education systems as they reopen As the crisis shifts from response to reopening,
following a crisis, which makes ongoing evaluation an these data can be used to inform policy responses
important source of knowledge to strengthen future aimed at mitigating impacts on schooling and
response. may be used to evaluate and adjust approaches.
The evidence on monitoring school reopening to sup- • Including communities in decisions about
port evaluation and adaptation suggests that: what is monitored as schools reopen, and
how, could have long-term impacts on school
• Data collected during the reopening phase
participation and achievement. It is crucial to
(from education or other sectors) can be used
decide early on what data will be needed to mon-
to adapt implementation approaches and for
itor reopening, and how they will be collected
longer-term planning. Several initiatives have
and used to monitor progress and, later, to eval-
begun to gather data during the COVID-19 cri-
uate the process.
sis, and may also be used to monitor reopening.
Targeting Resources Where They Are Most Needed 53. Getting Children Back to School
Students, teachers, and households are facing new
pressures on their time and resources that will make
Recommendations
reenrolment challenging for some families as schools
reopen. Governments should implement univer- To encourage and support the reenrolment of
sal campaigns to encourage enrolment and consider all students, policymakers should:
additional measures to support the transition back to
• Combine community participation and
school, including cash transfers and school meals tar-
large-scale direct communication cam-
geted to the most vulnerable.
paigns to parents, and consider increas-
Policymakers will want to access evidence to support ing attendance options to accommodate
their planning and decision making and to draw on all children, including those with high-
relevant experience from elsewhere. This brief sum- est risk of dropping out.
marizes the available rigorous evidence related to
• Provide financial or in-kind support,
getting children back to school and identifies two rec-
such as school feeding, to help families
ommended actions for policymakers based on that
overcome the increased costs of attend-
body of evidence.
ing school.
Summary of Evidence
Encouraging reenrolment may be needed to reenrol students who have new
Very little rigorous evidence exists on student reenrol- demands on their time.
ment and school recovery after a crisis. This is an area
There is also a large body of evaluating evidence from
where current monitoring could greatly contribute to
nonemergency settings that provides insight into how
future knowledge.
to use information campaigns and community partic-
Available evidence suggests that governments can ipation to increase enrolment. Much of this evidence
identify and prioritize students who are the most likely remains relevant to reopening and recovery after
to drop out by using existing data on risk factors, such COVID-19.
as precrisis attendance rates and relative economic
Evidence on encouraging reenrolment after a crisis
disadvantage. In Sierra Leone, an intervention to pro-
indicates that:
vide protective spaces during the Ebola crisis allowed
young girls to allocate time away from men, prevent- • Community participation is at the heart of
ing out-of-wedlock pregnancies and enabling them to disaster recovery and can support reenrol-
reenrol in school postcrisis. The program of safe spaces ment efforts. Effective strategies include financ-
and skills training almost entirely reversed the large ing parent-teacher associations, working with
(16 percentage point) dropout rates among adolescent local political and school management groups,
girls without access to the program. Other experience and creating opportunities for community
from Sierra Leone suggests that flexible approaches monitoring.
6• Large-scale direct communication with stu- Overcoming increased costs of school
dents and parents can also increase enrolment attendance
and attendance. Proven interventions in this Rigorous evidence suggests that financial support to
category include sending reminders and nudges vulnerable households is critical to getting children
via SMS, information campaigns to communi- back to school. Recovery efforts must help families
cate the benefits of education, and working with overcome new financial constraints and the oppor-
role models to share education success stories. tunity costs of schooling. Strategies could include
school-related fee reductions, subsidies, and house-
• Administrative approaches to increasing flex-
hold cash transfers.
ibility, such as double-shifting, are common
in crisis-affected contexts, but their impacts According to data from Liberia’s high-frequency phone
are largely unknown. Despite its popularity, survey, initiated during the Ebola epidemic, one in four
there is no rigorous evidence on the effective- households with children over the age of 12 reported
ness of double-shifting in crisis settings. Evi- that one month after reopening, their children had not
dence from stable low-income countries shows returned to school; and reenrolment rates were lower
that it can increase access and enrolment—and among girls than boys. Eighty percent of these house-
that it may increase opportunities for formerly holds cited a lack of money as the primary reason for
out-of-school-children to enrol once the schools their child’s nonreturn.
reopen. Over the longer term, however, a reduc-
tion in contact hours may reduce learning, Income shocks to poor households may induce parents
particularly in the poorest schools. Other prom- to take their children out of school and send them to
ising changes to the school environment include work. When schools were closed during Sierra Leone’s
improving school WASH (water, sanitation, and Ebola outbreak, many of the approximately 3 million
hygiene), which has been shown to increase children living in communities affected by the disease
enrolment and attendance, particularly among carried out petty trade or other forms of work to sup-
girls, although this is not always the case. port the survival of their households. Once withdrawn
from school due to parental unemployment or disas-
• Newly vulnerable children who are not able ters, many children will not subsequently reenrol.
to return to regular school environments will While measures should be taken to reduce the need for
still need opportunities to enrol and learn. children to work, it is possible that some students will
Some children may be at greater risk and danger need to continue working while they attend school,
of infection during this current pandemic, so and this will require flexible solutions.
some families may not be able to send their child-
ren back to school. Very little evidence yet exists Evidence regarding providing support to households
on the effects of COVID-19 on children with in overcoming the costs of schooling shows that:
comorbidities, such as preexisting respiratory
• Financial support, including cash transfers,
illness or immune dysfunction. Households with
can reduce the financial burden and opportu-
affected children—or in which other family
nity costs faced by families sending their chil-
members are at significant risk—may choose to
dren back to school. Intervention findings in
avoid schools even if they reopen. Reenrolment
noncrisis settings suggest that transfers to poor
plans should find locally relevant means of pro-
households in underserved regions offer the
viding opportunities to learn for these newly
highest return, especially where labour mar-
vulnerable students.
ket opportunities for children are likely to draw
them away from school.
Getting Children Back to School 7• School meals can encourage reenrolment and influenced reenrolment, but extensive litera-
attendance. Many governments, including ture supports the removal of fee barriers as a
some in collaboration with the World Food Pro- means of increasing enrolment and retention,
gramme, have arranged to continue providing with some studies showing larger enrolment
school meals while the schools are closed. The benefits among students considered to be at risk
shock associated with the crisis could increase of withdrawing. Where fees are levied—which is
the numbers of those in need of such resources. common among secondary schools in many con-
Offering school meals and basic necessity kits texts—social transfers have been used to reduce
could incentivize and make reenrolment possi- financial constraints and increase enrolment.
ble for some, particularly the most vulnerable. The removal of other costs, such as for uniforms,
have also shown strong positive impacts on
• Fee waivers may support reenrolment. Such
school participation in several countries.
waivers were used in post-Ebola Sierra Leone.
There is no rigorous evidence on how they
8 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for Policymakers4. Making School Environments Safe
Maintaining the health and safety of people and envi-
ronments will be more important than ever before
Recommendations
in the aftermath of the COVID-19 school closures. To
make school environments safe, additional health and To provide safe school environments follow-
hygiene measures should be implemented, and school- ing a pandemic, policymakers should:
based psychosocial and nutritional support should be
• Pair school-based hygiene promotion
extended to students to strengthen their overall health
with the distribution of waterless hand
and well-being in the wake of the pandemic.
sanitizer and/or soap (where hand-
Policymakers will want to access evidence to support washing stations are already available).
their planning and decision-making and to draw on
• Consider school-based screening for
relevant experience from elsewhere. This brief sum-
fever and cough, which may reduce
marizes the available rigorous evidence related to
risk and improve confidence, but
making school environments safe and identifies four
which does not by itself offer a reliable
recommended actions for policymakers based on that
solution.
body of evidence.
• Train and support teachers and other
school staff to offer school-based psy-
Summary of Evidence
chosocial support to returning students.
Health and safety measures
• Prepare for a spike in the number of
Priorities for education-focused efforts after the coro-
students with malnutrition and other
navirus pandemic should include ensuring that school
unmet basic needs.
infrastructure and facilities are safe, that rooms are
clean and disease-free, and that staff are prepared to
prevent the spread of disease in the event of a future
outbreak. Evidence on good school health and safety after a crisis
shows that:
Evidence on the impacts of improving school-based
health and safety measures remains thin; it should • Distribution of alcohol-based hand sanitizer is
be strengthened as schools around the world reopen. likely to offer an effective, scalable solution for
Existing evidence points to the importance of making hand hygiene for the prevention of COVID-19
sure that schools are equipped with adequate health transmission, especially where running water
and hygiene infrastructure and monitoring capabil- is limited. A substantial body of evidence, largely
ities, that students and staff practice regular hand- from wealthy countries, shows that young chil-
washing, that schools have access to sufficient hand dren in particular are more likely to use water-
sanitizer, and that students and staff are continuously less hand sanitizer consistently, when available,
screened for temperature and symptoms. than they are to use soap and water. Programs
9that supply hand sanitizer for use in classrooms due to the prevalence of mild or asymptomatic
and day care centers have recorded substantial coronavirus in children and the high baseline
reductions in diarrheal and respiratory illnesses. prevalence of cough and fever among young
Alcohol-based hand sanitizer is highly effec- children. Symptomatic screenings should com-
tive against COVID-19 and recommended for plement rather than substitute for hygiene and
use by the WHO. Compared with major WASH distancing measures.
(water, sanitation, and hygiene) infrastructure
upgrades, it is easy to distribute and deploy. Child protection and well-being
Crises and prolonged school closures can negatively
• Despite a mixed track record on water and impact the well-being of children and adolescents. Evi-
sanitation interventions in schools, access to dence suggests that armed conflict and environmental
hygiene supplies and handwashing promo- disasters have multiple negative impacts on children
tion can help prevent the transmission of com- and their families, including the potential to increase
mon infectious agents. Globally, school-based children’s engagement in dangerous forms of work.
handwashing programs have been inconsistent
in preventing common childhood conditions, Evidence from the Ebola crisis suggests that women and
such as diarrhea, respiratory infections, and girls were at a disproportionate risk of sexual exploita-
soil-transmitted helminths. WASH interventions tion and early marriage as a result of prolonged school
appear to be more effective when they are well closures. Household shocks resulting from the pan-
implemented, when they effectively target the demic may also lead to higher rates of child malnutri-
sources of disease transmission, and when they tion and difficulty meeting basic needs.
combine supply-side measures with behaviour-
change approaches. The evidence on child protection and well-being after
a crisis shows that:
• School-based temperature screening has
been a part of containment for previous out- • Psychosocial and emotional support are often
breaks, but its effectiveness against COVID-19 needed in schools following crises. Evidence
remains uncertain. Following the Ebola-related from studies of conflict, natural disasters, pub-
closures of 2014–2015, school reopening proto- lic health crises, and other humanitarian cri-
cols included daily temperature checks at the ses show that a child’s psychosocial well-being
school gate, but it is unclear whether these mea- can be negatively impacted by shocks. A rapid
sures reduced transmission. During the SARS systematic review of the impact of social isola-
epidemic, Singapore issued a thermometer to tion suggests that children are likely to experi-
every student and mandated twice-daily tem- ence high rates of depression and anxiety during
perature checks. This helped restore confidence social isolation and for several years after it ends.
among parents and students but did not lead to Evidence from past humanitarian crises and
any identification of cases. School systems could situations of displacement finds that teachers
consider screening children and school staff for trained in programs that prioritize learning and
temperature and cough—the two most common social-emotional well-being can help improve
COVID-19 symptoms, but such an approach learning outcomes and student well-being fol-
would probably have low sensitivity (potentially lowing a crisis. The broader education work-
missing infected children) and low specificity force, including nurses and counsellors, can
(potentially excluding noninfected children) support children as they return to schools.
10 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for Policymakers• Differential support may be needed accord- not having enough to eat; in the aftermath, acute
ing to gender and age. Following the Ebola epi- malnutrition spiked significantly. The World
demic, evidence suggests that adolescent girls Food Programme warns that the number of peo-
faced disproportionate risks and abuse while the ple at risk of acute hunger may double by the end
schools were closed. Interventions focused on of the year due to the coronavirus crisis. Surveys
the empowerment and well-being of girls may suggest that many families are already eating
be needed as schools reopen.
less and may quickly exhaust their household
• Schools should prepare for a spike in students resources. Quick and safe restoration of school
with malnutrition. During the Ebola epidemic, feeding programs is essential to help stem and
76 percent of children in Sierra Leone reported reverse nutritional losses.
Making School Environments Safe 115. Recovering Learning Loss
To address learning loss, policymakers should con-
sider targeted programs for accelerated recovery and
Recommendations
use low-cost coaching and communication methods to
support teachers and engage parents. To recover learning loss following a pandemic,
policymakers should:
Policymakers will want to access evidence to support
their planning and decision making and to draw on rele- • Engage students in accelerated learning
vant experience from elsewhere. This brief summarizes interventions to reverse crisis-related
the available rigorous evidence related to recovering learning loss and strengthen future
learning loss and identifies three recommended actions learning trends.
for policymakers based on that body of evidence.
• Engage teachers in training and coach-
ing so they can help students catch up,
Summary of Evidence and ensure that school environments
are safe and protected.
Accelerated learning
Decades of research shows that school closures are asso- • Engage parents by capitalizing on their
ciated with learning loss, with more pronounced effects current involvement in remote learning
among low-income families. Accelerated learning pro- to improve future outcomes.
grams are common interventions to support catch-up
efforts; they help students cover core academic mate-
rial in less time. Evidence suggests that accelerated cohorts. While not rigorously evaluated, Sierra
learning interventions focused on the basics can have Leone’s experience with simplifying and acceler-
positive impacts on learning in developing countries, ating a planned curriculum to cover two academic
particularly among those furthest behind. years in one following the Ebola pandemic sug-
gests that accelerated learning can help get stu-
As policymakers take steps to reverse learning loss
dents back on track. Variations of this approach
after the current crisis, they should keep in sight the
have shown positive results in other countries
social and emotional needs of children. Social isola-
as well. The idea is not to do more with less time
tion and loneliness are harmful to the mental health of
but rather facilitate quality catch-up with a sim-
children. In addition to accelerated learning, research-
plified curriculum of select core components to
ers are urging governments to allow time for play and
cover thoroughly in the time available.
socialization as children return to schools.
• Targeted, intensive programs focused on basic
The evidence on recovering learning loss after a crisis
literacy and numeracy can support acceler-
shows that:
ated learning, particularly for those furthest
• Simplifying the planned curriculum to acceler- behind. Following prolonged closures, students
ate learning can support recovery when learn- will return to school with varying levels of learn-
ing loss is likely to have occurred for entire ing loss, possibly due to previous skill levels,
12differential access to distance learning, or varying Supporting teachers and parents
home support during the closures. A robust body Evidence from past crises suggests that teachers need
of evidence indicates that instruction focused on new skills and capacities as they support student transi-
foundational literacy and numeracy skills tar- tions back to school, ensure safe school environments,
geted at students’ individual levels, even for one and provide remedial learning activities. Not only does
or two hours per day, can substantially improve providing support to teachers offer the potential of
learning. This kind of targeting can be based on improved learning, such efforts have also been linked
a rapid performance assessment as students to the improved well-being of students. While there
return. are many promising practices to strengthen teacher
development over the long-term, we offer evidence
• Teacher-led learning camps and remedial
on activities to support teachers and classrooms in the
tutoring programs can support learning accel-
immediate-term after schools reopen.
eration. Evidence suggests that learning camps
prior to and between school terms can support The pandemic has also forced the increased involve-
student learning. Such camps provide focused ment of parents around the world in their children’s
opportunities to accelerate the mastery of education while schools are out of session. This pro-
basic skills. Carefully planned programs using vides a unique opportunity for encouraging parents
research-based instruction can be highly effec- to remain engaged over the longer term even as the
tive if students consistently attend. Studies find schools reopen.
positive impacts of high-quality remedial tutor-
The evidence on supporting teachers and parents after
ing for marginalized students who have missed
a crisis suggests that:
prolonged periods of schooling due to conflict or
displacement. • Regular coaching can support teachers as they
adapt to new demands. As schools reopen,
• Where connectivity allows, remote- and tech-
teachers will face the daunting task of support-
nology-based learning strategies can also sup-
ing students in covering lost ground. They will
port remedial and accelerated learning. Many
be operating in a modified environment and
countries around the world have made impres-
will likely be delivering an irregular syllabus. All
sive progress in developing distance learning
of this requires new skills and capacities. For-
platforms during school closures. When schools
mal retraining will take time and substantial
reopen, these platforms could complement reg-
resources. Evidence suggests that on-the-job
ular instruction to support learning recovery.
and virtual teacher coaching can help strengthen
For example, evidence suggests that high-quality
teaching practices quickly and effectively.
adaptive-technology-aided after-school instruc-
tion programs can accelerate learning growth, • Leveraging the entire education workforce—
particularly for those furthest behind. How- not only the teachers—may support learning.
ever, while such remedial strategies can be use- Evidence suggests that engaging other education
ful, they are not all the same. Simply providing professionals, volunteers, and peer-to-peer sup-
hardware—such as computers or tablets—with- port as tutors, mentors, and additional help for
out carefully designed software is unlikely to teachers can strengthen learning outcomes as
improve learning outcomes. well as support vulnerable students, including
girls, in returning to and staying in school.
Recovering Learning Loss and Building Back Better 13• There are several low-tech and low-cost options in large learning gains for their children. Past
to maintain high rates of parental engagement interventions include informing parents of their
in the schooling of children. Direct interven- children’s learning levels, progress, and atten-
tions with parents via SMS or other channels dance. Where feasible, personalized messages
has been shown to increase and sustain their may be more effective in strengthening parental
involvement at home and at school, resulting engagement than generic messages.
14 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for PolicymakersTechnical Appendices
Evidence Appendix 1.
Engaging Communities in School Reopening Plans
Trust, risk, and community participation SARS, and evidence from qualitative work suggests
Without trust in government, support for reforms is that similar activities also supported recovery in Sierra
difficult to mobilize, particularly when short-term sac- Leone after Ebola.
rifices are involved. Low levels of trust and widespread
Evidence from the Ebola epidemic suggests that a
concern about misinformation have been associated
primary reliance on big data absent substantial com-
with a decreased likelihood of following Ebola guide-
munity involvement in case reporting and recovery
lines in the Democratic Republic of Congo. In Sierra
planning slowed containment and the subsequent
Leone, trust and community involvement led to an
recovery.
increase in community monitoring and responsive-
ness to government policies, as well as perceived qual- Strategies that leverage local social networks are bet-
ity and use of services. ter able to draw on community information, tools, and
assistance to support faster recovery; and feedback
Perceived risk, more than actual risk, determines
mechanisms allow affected communities to express
a population’s reaction to a crisis. Any mismatch
their opinions about the distribution of resources,
between risk perception and actual hazard can reduce
providing data to inform choices and increase
compliance with government plans during recovery,
accountability.
including returning to school. Information that is con-
flicting or that underestimates the severity of contin- Communication strategies
ued risk may lead to confusion or noncompliance with There is limited evaluation evidence on the impact
recovery plans. When information is provided about a of specific communication methods during a crisis,
crisis and its related risks, risk perception among par- which likely varies across contexts. Source credibility
ents and school staff can quickly shift. is critical for effective response and recovery in an era
of mass communication and misinformation.
Recovery efforts will also be impacted by changes in
risk perception during the response. Before COVID- Internet connectivity is weak in many areas, but social
19, SARS was the leading example of the widespread media and SMS can open new avenues for improving
use of nonmedical public health measures to contain strategic communications across various stages of a
an infectious disease outbreak (e.g., social distancing). crisis, including the promotion of the public’s com-
The use of such measures impacted the public percep- mon responsibility and cooperation. Mobile and radio
tion of risk—on one hand heightening the perception connectivity is more widespread and could serve as a
of serious risk of infection and on the other providing practical way of providing information to vulnerable
reassurance that actions were being taken to reduce groups when face-to-face measures are not feasible.
that risk. Highly visible public health activities, such as Telephone information campaigns have been effective
temperature readings for children and wearing masks, in relieving anxiety and improving knowledge about
supported response and recovery in China during a crisis among vulnerable adults. The gray literature
Technical Appendices 15strongly underlines the central role of local communi- campaign paired with a telephone hotline positively
ties in the dissemination of information. The choice of impacted beliefs and behaviours following Hurricane
communication channels should ultimately be deter- Katrina in the United States. In Singapore, an effective
mined by local resources and communications norms. integrated communication campaign during the SARS
Utilizing national household surveys to determine crisis included television, press conferences, adver-
household access to various forms of communication tisement, booklets, posters, and a public hotline for
technology is a good place to start. anyone seeking information and for providing feed-
back. Mixed-media campaigns have also been used
Combining multiple forms of media likely sup- to encourage school enrolment programs and health
ports communication efforts. For example, a media care compliance in nonemergency settings.
Evidence Appendix 2.
Targeting Resources Where They Are Most Needed
Coordinating actions and resources
While this is a challenging area in which to generate less educated households were less prepared for disas-
rigorous evidence, the disaster response literature ters and faced greater challenges during recovery.
finds evidence that strategic coordination and clear
communication among sectors and agencies improves Substantial evidence from social transfers, including
implementation of response and recovery plans. A those targeting children, suggests that targeting is key
study examining disaster response and recovery fol- to intervention effectiveness, a lesson relevant to crisis
lowing a major earthquake in Haiti found that collab- recovery as well. The effectiveness of targeting is shaped
oration increases exchange and can create incentives by program design features, including group eligibility,
that strengthen recovery. The findings suggest that location selection, and relative allocation of benefits. A
organizations benefit from pooling knowledge and quantitative study comparing targeting mechanisms in
data and from sharing best practices; leveraging social transfer programs in sub-Saharan Africa found
comparative advantage and economies of scale; and community targeting to be an effective tool for targeting
expanding reach while avoiding duplication of efforts. those who would most benefit from the program.
Targeting and inclusion While it is critical to target the most vulnerable during
Disasters do not impact all groups in the same way. The and after crises, evidence from the literature also cau-
poorest, women, those with disabilities, and otherwise tions that crises may increase the number of poor and
disadvantaged groups often experience dispropor- can lead to new vulnerable groups. A study in Bangla-
tionate negative impacts during and in the aftermath desh tracking child malnutrition following a severe
of disasters. A study examining school attendance fol- three-month flood found evidence of a “crossover”
lowing Hurricane Katrina in the United States found phenomenon in which the nutritional status of tar-
that schools with a higher proportion of economically geted children improved while malnutrition increased
disadvantaged students had more difficulty recovering among some who were not initially targeted. This sug-
following the crisis. A study on disaster preparedness gests that it is important to assess changes in vulner-
and recovery in the Philippines found that poorer and ability resulting from a crisis and that recovery may
require broader targeting than an initial emergency
response.
16 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for PolicymakersMonitoring and adapting of Ebola through a monthly phone survey, providing
It is likely that traditional data collection methods— data on schooling and broader economic impacts that
school census and large-scale face-to-face surveys—are might affect recovery.
disrupted by the crisis. Using mobile phones to conduct
Evidence from Uganda suggests that creating oppor-
short and frequent surveys can produce high-quality
tunities for community engagement, allowing partici-
data more quickly—and more cheaply on a per survey
pants to influence how and what they will monitor and
basis—than traditional methods. Mobile phone surveys
manage, delivers strong impacts on achievement and
were found to be valuable tools for crisis monitoring
absenteeism.
during the Ebola epidemic in Liberia and Sierra Leone.
Liberia’s Institute of Statistics monitored the impacts
Evidence Appendix 3.
Getting Children Back to School
Encouraging reenrolment own success story with a household. Other effective
Carefully designed approaches to strengthen school– approaches include providing information to parents
community linkages have had positive and substantial on their children’s absenteeism—which may also pos-
impacts on attendance. Leveraging parent associations itively impact siblings—and using high-frequency SMS
and motivating participation by involving parents in to communicate with parents in real time regarding
the management of school grants has improved enrol- student attendance.
ment and retention in poor communities. Engaging
with local political structures may reinforce existing Overcoming increased costs of
school-based management arrangements. Another school attendance
promising example of local-level engagement is com- Conditional cash transfer programs in which school
munity monitoring of service levels and provider attendance is the condition has been shown to increase
responsibilities, which improves health service utiliza- educational achievements and mitigate the negative
tion, quality of service delivery, and outcomes. impacts of household shocks. However, increases in
Examples of effective enrolment and attendance rates of educational attendance may occur without an
strengthening include improving perceptions among equivalent decline in the incidence of child labor and
students of the benefits of an education by provid- mainly at the expense of child leisure. In some emer-
ing them with information on its returns, or hav- gency recovery phases, access to credit has been shown
ing an appropriate role model describe the benefits to diminish the immediate financial burden on house-
of an education, such as a person sharing his or her holds, reducing the need for child labor.
Technical Appendices 17Evidence Appendix 4.
Making School Environments Safe
Health and safety measures significantly decreased rates of diarrhea and vis-
There is no rigorous evidence that we know of on the its to health clinics among the younger siblings
impact of WASH interventions in schools following of students.
a crisis. Some evidence suggests that outbreaks may
• A multicountry program across primary schools
lead to the long-term adoption of healthy habits, such
in Cambodia, Indonesia, and Laos, which
as handwashing, even after the immediate crisis has
included daily handwashing with soap as a
passed. With appropriate recovery planning, it may
group activity in addition to a deworming treat-
be possible to evaluate the effectiveness of the various
ment, did not reduce parasitic infections or the
health and safety measures that are being used during
prevalence of underweight, although this may
the current crisis.
be attributable to an unexpectedly low parasitic
In nonemergency settings, WASH interventions in prevalence at baseline.
schools show a mixed track record of implementing
• In rural China, the promotion of handwashing,
preventative measures including handwashing sup-
the provision of soap, and the training of peer
plies or efforts to change handwashing behaviour—
hygiene monitors reduced illness-related school
the component of WASH most relevant for COVID-19
absences by more than 50 percent.
transmission.
• In Cairo, children required to wash their hands
• In Laos, where primary schools were upgraded
twice a day (a behaviour reinforced with health
with running water, handwashing stations, and
messages) saw major reductions in influenza-like
other upgrades, and following a behaviour-
illness (40 percent), laboratory-confirmed influ-
change campaign, there were no significant
enza (50 percent), diarrhea (30 percent), and
recorded improvements in rates of diarrhea,
conjunctivitis (67 percent).
respiratory infection, conjunctivitis, or
soil-transmitted helminthiasis (STH) infections. • In Mali, a multifaceted school WASH interven-
tion was associated with decreased rates of diar-
• In Kenya, a multifaceted school-based interven-
rhea and lower respiratory infections. Greater
tion—which included water treatment, water
adherence to the handwashing component
containers for hand washing and drinking
was associated with decreased rates of school
water, latrine construction, and hygiene promo-
absence.
tion—helped reduce reinfection with one hel-
minth among girls but had no impact on three • In Bihar, a school-based behaviour change pro-
other STH infections. The intervention substan- gram to encourage handwashing with soap pro-
tially reduced diarrhea among children without duced only modest increases (4.4 percentage
access to drinking water from other improved points) in the prevalence of handwashing with
sources but did not reduce it among those who soap.
could access an improved water supply within
one kilometer, which may suggest that clean • In Peru, school-based handwashing promotion
drinking water rather than handwashing drove improved handwashing behaviour but had no
the observed gains. The intervention also led to impact on evaluated health outcomes.
18 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for PolicymakersA substantial body of evidence suggests that alco- • Following the Ebola-related closures of 2014–
hol-based waterless hand sanitizer is a highly effective 2015, school reopening protocols in Sierra Leone,
substitute for handwashing, particularly for younger Guinea, and Liberia included daily temperature
children who may find it difficult to adhere to proper checks at the school gate plus soap/chlorine dis-
handwashing techniques. Although such hand sanitiz- tribution and installation of handwashing sta-
ers are not as effective as soap and water under per- tions. As of August 2015, UNICEF reported zero
fect-use conditions, particularly against some bacterial known instances of a teacher or student becom-
agents, the World Health Organization and the Centers ing infected with Ebola at a school. It is unclear
for Disease Control and Prevention recommend their whether transmission would have occurred in
use against COVID-19, and studies have confirmed the absence of these measures.
their high levels of efficacy.
• During the SARS epidemic, Singapore issued
• In Nairobi’s primary schools, the availability of thermometers to every student and mandated
waterless hand sanitizer was associated with twice-daily temperature screening for all stu-
reduced prevalence of rhinorrhea (but not with dents; any child with an elevated temperature
statistically significant declines in gastrointesti- was excluded from school. Students with con-
nal or respiratory illnesses). Children were more sistently high temperatures were referred to a
likely to clean their hands after using the toilet health center for further examination. Tempera-
if hand sanitizer was available (compared with ture screening did not result in any case identifi-
soap and water). cation but reportedly did help reassure students
and parents of school safety.
• Substantial evidence from wealthier countries
shows young children with access to waterless • Symptomatic screening for COVID-19 would
hand sanitizer are more likely to clean their likely focus on observed temperatures and pres-
hands—and have substantially lower risk of diar- ence of a cough. However, this is likely to offer
rheal and respiratory disease and related absen- a crude screening tool at best, excluding many
teeism—than children with access to soap and children without coronavirus infection while
water alone. missing others who are asymptomatic. Most
coronavirus infections in children are mild, and
The evidence base for school-based management
many are asymptomatic—that is, they would
of outbreaks also remains thin. Use of thermome-
not be detected based on symptomatic screen-
ter-based screening coupled with teacher training on
ing criteria. We do not yet know the extent of
the management of suspected cases may allow teach-
asymptomatic transmission in children, but
ers to detect fever in children upon school entry and
presymptomatic transmission in adults can be
exclude ill children from entry. Schools could also
extensive. On the flipside, mild coronavirus
consider prohibiting entry to a child with a cough.
symptoms may be indistinguishable from com-
Evidence on the utility of a symptomatic screen-
mon childhood ailments. In children under five,
ing-and-exclusion approach is limited and compli-
Demographic and Health Surveys across 182 low-
cated by the relatively high prevalence (and potential
and middle-income countries (conducted
transmissibility) of asymptomatic (or presymptomatic)
between 1986 and 2012) suggest a 38.4 percent
COVID-19 infection in children, as well as widespread
average prevalence of either cough or fever
endemicity of other respiratory and febrile diseases.
within the two-week period prior to the survey
Technical Appendices 19date and an 11.7 percent average prevalence of In some contexts, girls may require additional sup-
both symptoms over the same period. These fig- port following a crisis. High-quality girls empower-
ures are not directly applicable to school-going ment and livelihood programs have been shown to
cohorts (e.g., older children)—and are likely out have strong positive effects on multiple measures of
of date—but nonetheless, they point to the diffi- girls’ empowerment, including persistence in school;
culty of distinguishing mild COVID-19 symptoms school-to-work transitions; a reduction in early mar-
from other common endemic infections. riage and risky behaviours; improved networks, aspi-
rations for the future; and in some cases, parental
Child protection and well-being beliefs about education for girls, particularly in unsta-
Evidence from studies of conflict, natural disasters, ble contexts. Synthesis results suggest that high-quality
public health, and other humanitarian crises show that girls’ empowerment programs can: (1) be implemented
shocks can negatively impact psychosocial well-being in environments prone to instability; (2) can shield
of children. School-based interventions and school- girls’ from the worst impacts of a crisis during a coun-
based referral models are encouraged to help students try-wide shock; and (3) are cost-effective.
recover. Evidence from humanitarian crises shows
that supporting teachers during reconstruction can Acute malnutrition can spike during a crisis. During
improve the psychosocial well-being of students. Inter- the Ebola crisis, 76 percent of children in Sierra Leone
ventions such as the International Rescue Commit- reported not having enough food to eat; after the crisis,
tee’s “Healing Classrooms” have had positive impacts rates of severe acute malnutrition rose significantly.
on children’s social emotional well-being following
Initial signs suggest COVID-19 lockdowns have already
a crisis across contexts. Following the Katrina hurri-
had a severe effect on food security. The World Food
cane, cognitive behavioural therapy and self-reporting
Programme estimates that the number of people
helped reduce posttraumatic effects. Using available
at risk of acute hunger may double by end-2020 as a
resources within schools has shown positive effects.
result of the crisis. BRAC surveys across eight African
The broader education workforce, including school
and Asian countries show that many households have
nurses and counselors, can also support the well-be-
less than one week of food stocked (up to 90 percent
ing of children returning to school. In Japan, for exam-
of households in Sierra Leone); in Liberia, 84 per-
ple, support from teachers, nurses, and counselors
cent of households report that food consumption has
improved the long-term mental health of students.
dropped “a lot” since the start of the crisis. In Bangla-
As with other forms of crises and postcrises support, desh, 14 percent of respondents report having no food
targeting is important. A systematic review of 24 stud- in their homes, and almost half of families have less
ies on mental health and psychosocial interventions in than three days of food. Prompt and complete resto-
conflict-affected areas highlights the predominance ration of school feeding programs is crucial to blunt
of school-based interventions to support students nutritional losses and prevent further declines.
returning to school following a crisis. Despite overall
positive effects, the evidence suggests that the benefits
of such school-based interventions tend to be limited
to subgroups of children.
20 Planning for School Reopening and Recovery After COVID-19: An Evidence Kit for PolicymakersYou can also read
