Latest evidence on the indirect impacts of C19 on education and learning — 5 January 2021

Rachel Glennerster, Chief Economist, Sarah Lane Smith and Hannah Chisambi, Education Research Team, and Anna Rudge, Senior Economist, Foreign, Commonwealth & Development Office (FCDO).

As we move into 2021, we will continue to look at rigorous evidence that shines a light on the many indirect impacts of Covid-19 on peoples’ lives globally. In May last year, we saw new research showing the long term impact of learning as a result of school closures during the Pakistan Earthquakes by RISE. The new research summarized below, often supported by FCDO, talks to the impacts from Covid-19 measures already seen on education and learning globally. For example, 16% of the surveyed 19-year-olds in formal education in Peru have dropped out, and for pre-primary children in Ethiopia half of respondents reported that they had no children’s books at home.

Context

1.6 billion young people were impacted by school closures at the initial peak of the COVID19 pandemic response in 2020 — and the World Bank estimates circa 670,000 still face full school closures. With schools not yet reopened globally, the full impact of the COVID-19 shock to education is yet to be understood — but is expected to be large. Simulations show learning loss far beyond that of the equivalent learning that would have been achieved in the school closure period, and the OECD is predicting that a loss of one-third of a year of schooling will lower a country’s GDP by an average of 1.5% over the remainder of the century. Increasingly, evidence shows that the crisis has impacted marginalized learners the most. Decision makers need to consider quickly how best to mitigate learning loss and how to build resilient systems of education able to support children to catch up learning.

This unprecedented level of education disruption sits within a pre-pandemic context of the learning crisis, in which it was estimated that 53% of children in low- and middle-income countries could not read before the age of 10. Progress in raising learning outcomes in recent decades has been extremely slow. Data collected from 95 non-OECD countries suggests that it would take 102 years for these countries to reach OECD average learning outcome levels. With education financing expected to decrease within a declining economic climate, the risks to global and domestic education budgets are stark — the recently launched Save Our Future Campaign estimates a likely scenario of annual financing gaps for low and lower middle income countries of between $178 and $193 billion over the next 10 years. High-quality evidence is integral to supporting effective decision making for the COVID-19 education response within an increasingly challenging fiscal space. FCDO funded programmes are working to generate critical evidence, deliver insights to inform the education response and support partners to mitigate learning loss.

Below are a summary of some interesting research currently being shared from both the FCDO and its partners.

Research and data

1. Education disrupted in Ethiopia, Peru, Vietnam and India affecting disadvantaged youths without online access the most- Results from FCDO funded Young Lives at Work Covid Telephone Survey. New research from the Young Lives COVID-19 phone survey in Ethiopia, India, Peru, and Vietnam (~11,000 respondents) is painting a worrying picture of how the economic and social impact of Covid-19 may not only be halting progress made over the last two generations but may be reversing life chances and deepening inequality for many young people, hitting those living in poor communities hardest. Emerging findings suggest that many students may never return to the classroom, reversing opportunities for finding decent work and escaping poverty. In Peru 16% of 19 year old’s have dropped out of education mainly for cost of fees and lack of internet access. The digital divide is affecting the most disadvantaged youth (in Ethiopia <5% of students could access online learning during lockdown).
2. Low-tech learning solutions in Botswana — Angrist, Bergman, Brewster & Matsheng. A rapid trial conducted in Botswana provides some of the first experimental evidence on strategies to minimize the pandemic’s impact on learning. Early evidence of effectiveness was found for two low-technology interventions to substitute schooling during lockdowns. SMS text messages and direct phone calls were tested at 4 to 6 week intervals, resulting in cost-effective learning gains equivalent to a 52% reduction in innumeracy. Researchers also found increased parental engagement in their child’s education and more accurate parental perceptions of their child’s learning. The results have immediate policy relevance within COVID-19 educational policy and could have long-run implications for the role of technology and home-based support as complements to the traditional education system.
3. How to support students and learning process during India’s school closures — CGIAR IFPRI (supported by the FCDO Agriculture Research). New evidence from survey data on 20,000 adolescents in northern India to estimate the impact of closed schools on educational achievement provides insights for how targeted government policies can make a difference. This Blog summarises the insights, key findings include:

• Adolescents with high parental support were approx. 20% more likely to show age-appropriate reading and math skills.
• Gender equal attitudes resulted in adolescents being twice as likely to have age-appropriate learning levels.
• Adolescents that had experienced domestic violence were 20–30% less likely to be able to read or solve subtraction equations.

4. Learning lessons from the impact of the Ebola crises to protect human capital accumulation for girls in Sierra Leone — FCDO funded research from J-PAL affiliates Bandiera, Buehren, Goldstein, Rasul, Smurra. Experimental long-term evidence from an intervention during Ebola school interruptions gives key insights into protecting girls’ education during COVID-19. During Sierra Leone’s outbreak, girls were at great risk of long-lasting harm. Younger girls spent significantly more time with men, leading to a rise in teenage pregnancies and a 17% drop in school enrolment post epidemic. An Empowerment and Livelihood for Adolescents programme (ELA), was found to protect girls from adverse outcomes related to schooling interruption. Girls who lived in participating villages were more likely to avoid early pregnancy and return to school. Dropout rates halved and there was a 7% reduction in out-of-wedlock pregnancies. A long-term follow-up in 2019/20 shows persistent impacts of the intervention on the human capital accumulation of young girls, time they spend with men, and quality of partners matched with. The analysis has important implications for 2020 school closures, highlighting the protective and lasting role that safe spaces can provide for young women facing violence.

5. EdTech across Africa — The FCDO funded EdTech Hub and eLearning Africa share the results of a survey of 1650 educators and technology specialists in Africa: The Effect of COVID-19 on Education in Africa and Its Implications for the Use of Technology. School closures during the pandemic were seen as essential by 92% of respondents. Results show:

• lack of technology access is perceived as the biggest learning barrier during the school closures, with early childhood and primary level students seen as most likely to be disadvantaged.
• TV and radio are deemed to be the most important tech for maintaining learning in the primary years, with online learning seen as most important for secondary level.
• Longer-term, 50% of respondents see COVID-19 as providing new opportunities for education systems, and 83% believe curriculum should be adapted for more effective distance learning in future.

6. Pre-primary education in Ethiopia — Implications of COVID-19 for pre-primary education in Ethiopia: Perspectives of Parents and Caregivers, funded by FCDO through the World Bank’s Early Learning Partnership Systems Research Programme, provides results from phone surveys carried out with 480 parents/caregivers in Ethiopia. The results suggest a high level of learning disruption for pre-primary aged children due to limited access to relevant learning resources in the home, with over half reporting no children’s books at home — rising to 72% if parents are illiterate. Only half of respondents report support to learning activities for pre-primary children during school closures with urban families reporting higher engagement than rural. Results show the burden of educational support is falling to mothers (almost half) despite Mothers being more likely to be illiterate themselves. Around three quarters of respondents reported engaging more often in play with their child than pre-pandemic, however results show that those from poorer households are less likely to play often with their child. More data on household economic impacts and child wellbeing can be read at the full blog here.

1 Source: https://www.ukfiet.org/2020/implications-of-covid-19-for-pre-primary-education-in-ethiopia-perspectives-of-parents-and-caregivers/

7. Education policy responses in 149 countries — UNESCO UIS have analysed the results of the first two rounds of the Survey of National Education Responses conducted with Ministry of Education Officials in 149 countries (Round 2). Interesting headlines include

• 73% of countries had fully or partially reopened schools by September 2020
• countries reporting missing an average of 47 days of in-person school instruction during the closures (equivalent to approximately 1/4 of an academic year).
• 60% of lower-middle-income and 39% of low-income respondents, required teachers to continue working during school closures. More policy perspectives in the full report here.

8. Large scale data on learning in India — Annual Status of Education Report (ASER) 2020 Wave 1 results have now been released, from the first ever ASER survey to be conducted by phone due to the pandemic. The household survey reached 59251 children aged 5–16, with 8963 schools also providing data. Results provide a breadth of evidence about learning in India during the crisis. Data on learning material access including digital devices showed that 61.8% of households had smartphone access, with more than 80% of children having relevant textbooks at home — however state variation is evident with this figure dropping below 50% in some areas. Around 1/3 of children had received some form of additional learning materials in the week prior to the survey with, again, large state variation in this figure (Punjab — 87.6% of compared to West Bengal — 20.5% for example). Approx. 75% of children are receiving some kind of study support from family members, however Father engagement remains low in comparison to Mothers and older siblings.

Models and Simulations

9.Can we know the long-term learning losses? Kaffenberger & Pritchett (RISE). Using empirical data from previous studies, researchers from FCDO funded Research on Improving Systems of Education programme (RISE) are using an innovative methodology to model potential long-term learning loss from the temporary shock of Covid-19 related school closures. Weaknesses in existing education systems mean that, when and if many children return post-Covid, they will be even further behind the curriculum in a system which is not set up to effectively target learning at children’s level. The RISE model shows that children could lose more than a year of learning from just a three-month school closure, as short-term learning losses are compounded by children returning to systems that were already in crisis. There are things that can be done. Under a short-term remediation scenario where missed curriculum content is covered immediately when students return, the long-term adverse impact could be reduced to half a year’s worth of learning lost. RISE consider a scenario in which short-term remediation is combined with a longer-term reorientation of teaching to children’s actual learning levels. This scenario not only fully mitigates the effect of the shock but could in fact increase long-term learning by more than a full year’s worth of instruction. This work highlights the urgency of post-pandemic reform to address the learning crisis and to help education systems build back better.

2 Snapshot of RISE Simulation Tool -https://riseprogramme.org/tools/simulating-learning

10. Simulating the cost of lost pre-primary education: IDB’s Economic Costs of Pre-primary Program Reductions due to COVID-19 Pandemic simulates the discounted value of losses in future income for preprimary closures of 3, 6, and 12 months for 140 countries. Losses are calculated as an estimated percentage of GDP due to reduction in preschool participation — which vary by country depending on rates of pre-COVID pre-primary participation, most notably resulting in smaller losses for low-income countries in which pre-primary access was previously low. Percentage losses vary dependent on number of months closed, and country income status with simulations showing 5.89% of GDP lost for a 12 month closure in high-income countries, compared with 2.94% (6 months) and 1.47% (3 months). Lower middle income countries could lost 5.32% (12 months) 3.38% (6 months) and 1.69% (3 months). The full report includes simulations by country for those interested in this detail.

Other useful reads and resources

• UNICEF Innocenti has launched a new child-focused evidence database — the Children and COVID-19 Research Library.
• CGDev have some recent interesting commentaries on COVID19 and girls education, the economic impact of the pandemic on education financing, and teacher labour markets following shocks.
• This GPE-hosted blog shares thoughts on the use of Citizen led Assessments as a best approach for assessing Foundational Skills during and after the COVID19 pandemic.
• Cost effective approaches to improve global learning: What does the recent evidence tell us are the Smart buys form improving learning in LICs and MICs? Report published by the Global Education Evidence Advisory Panel.

Follow Rachel Glennerster on twitter @rglenner

Follow @FCDOResearch for the latest FCDO research news, jobs and funding.