School closures and learning loss in West Africa

Worldwide, school closures from COVID-19 have caused large learning losses. A study by the World Bank in June 2020 simulated the possible effects of COVID-19 school closures, estimating that in the ‘pessimistic scenario’ of 7 months of school closures, the typical student would lose 0.9 years of quality-adjusted schooling¹. In many countries actual school closures were much longer than 7 months, cumulatively leading to millions of years of lost schooling and over $10 trillion in lost future earnings².

Evidence is beginning to emerge showing the effects of school closures will be dramatically different for low- and middle-income countries than for high-income countries. On average, low- and middle-income countries have experienced longer school closures, with limited capacity to transition to remote learning³. This lost year of learning is deeply concerning, not just because of the educational progress students have missed out on, but also because of the increased risk of gender-based violence, teenage pregnancy, and child marriage. Our recent findings from surveys of parents and caregivers in Ghana, Sierra Leone, and Liberia underscore these risks and offer some insight into how policymakers can respond.

We partnered with Rising Academy Network (RAN) and Echidna Giving to better understand the effects of COVID-19 school closures on educational outcomes and gender effects in West African countries. This research was done in part to understand if COVID-19 would have a similar effect as the Ebola pandemic in West Africa, which disproportionately impacted girls — with girls dropping out of school at a higher rate than boys amid a surge in child marriage, child sex work, child labor, and teenage pregnancy⁴. To understand whether the effects of COVID-19 would be similarly disastrous, we surveyed over 3,000 caregivers of students from 103 primary and secondary schools. In this post, we report on the first round of data collection, conducted from January to April 2021, where we attempt to estimate the magnitude of learning loss in these three countries. Our next post will report the results from the second round of data collection, conducted from July to August 2021, where we will explore what policymakers and education stakeholders can do to address these challenges⁵.

Our research found that school closures have likely led to alarming learning losses. Caregivers reported drastically reduced hours spent on educational activities when schools were closed. On average, students spent 5.7 fewer hours per day on education-related activities during school closures. Of course, this difference is largely explained by the fact that students were not in school on weekdays. But what surprised us was that there was little home-based learning being substituted for the lack of in-school learning (see Figure 1). 

Unlike many students in Europe and the United States, most students in Ghana, Liberia, and Sierra Leone did not have the resources at home to compensate for the lack of in-school learning. Fewer than one in four households had internet connectivity on a smartphone, and fewer than one in eight households had internet access that could be used for remote learning⁶. Due to the lack of access to these platforms, education providers in settings with limited resources have innovated with low-technology solutions such as Rising on Air for radio, Young1ove and Dost for Delivery over basic handset. Yet even these technologies do not reach all students: nearly half of households in our sample do not have access to a smartphone, basic handset, or radio. Consequently, no matter the technology used, a significant proportion of households will continue to rely on physical materials, like worksheets and textbooks, to access any remote learning during school closures. 

Challenges with remote learning are compounded for girls. In addition to lacking access to Edtech, girls often face an unsafe environment at home. In our survey, caretakers were concerned that girls in their communities were at increased risk of physical and sexual violence during school closures. In this light, it is hardly surprising that re-enrollment rates for girls when schools reopened were 98% since the classroom offers a safe space that is conducive for learning⁷. 

Education providers were also concerned that school closures could lead to a spike in teenage pregnancy. While we only found five drop-outs in our sample associated with teenage pregnancy, it is possible that this number is understated since our results were self-reported and stigmatization around teenage pregnancy is prevalent. Caregivers in our surveys predominantly believed that pregnant girls should not be sent back to school⁸. In the past, in fact, Sierra Leone banned pregnant girls from coming to school – causing them to permanently drop out⁹. It took 5 years for this ban to be overturned¹⁰. 

The search for immediate and sustainable solutions is critical, given the trends we have begun to see in past data could worsen in the future. Many West African countries are currently weathering their 2nd or 3rd COVID-19 surge, and some like Liberia have responded with a second round of school closures. Future COVID-19 waves and further school closures are imminent given the low vaccination coverage in the region. 

From the first round of our data collection, we learned that school closures due to COVID-19 in West Africa have likely led to devastating learning losses. We have also learned that lack of access to the internet and distance learning technologies are a major possible factor contributing to learning loss. Going forward, education providers will need to:

• Invest in tools to enable at-home learning, and ensure they can adapt their content to the medium of instruction. 
• Include non-technological options as distant learning cannot be implemented using a one-size fits all approach, and must include non-technological options. 

Given emerging evidence that even low-technology Edtech solutions fail to reach a majority of students, the solutions required must be out of the box¹¹. Our second round of data collection (currently underway) explores what these options could look like and will inform programming to support remote learning.