It is time to return to learning

This page in:
It has now been a year since schools started closing all around the world to try to contain the spread of COVID-19. It has now been a year since schools started closing all around the world to try to contain the spread of COVID-19.

For most students, the COVID-19 pandemic has been a tragedy of multiple dimensions. With schools closed, many children lost access to a relatively safe environment at a time when the deteriorated economic conditions at home led to increased stress as well as domestic violence. Meanwhile, 350 million of the poorest children in the world lost their main meal when schools closed. Given the huge digital divide, most countries attempted different forms of remote learning with platforms that combined online methods with TV, radio, and the distribution of learning materials. But no app, algorithm, or TV program can be a substitute for the teacher in the learning process or for students’ interaction in building social skills. The learning opportunities provided by schools to children in poor households are perhaps one of their few options for escaping poverty. In shutting down schools, societies were also shutting down one of the few social levelers.   

But closing down schools goes beyond losing core skills – there is mounting evidence of its effects on anxiety, depression, and considerable harm to children’s mental health and well-being. School closures have also been associated with an increase in suicide rates among children and adolescents. In addition, school closures reduce female labor participation and increase the gender wage gap.  

It has now been a year since schools started closing all around the world to try to contain the spread of COVID-19. At the outbreak of the pandemic, closing schools to protect children, teachers, and parents from a new unknown virus was the right thing to do. But after more than a year, enough evidence has been generated to take a more informed decision. Two primary elements should be at the center of the decision-making: the health risks associated with opening schools versus the education and learning costs of keeping schools closed.

We try to summarize here what we know about these two factors. Initial evidence comes from Europe and the US, where schools have been relatively quick to reopen after the first episodes of closings during the first wave of the pandemic.

The first concern of reopening schools is the potential risk for children’s health. But a study by the Center for Disease Prevention and Control in Europe found that most girls and boys under the age of 12 who acquire COVID-19 have no symptoms and a very small proportion have severe symptoms. The second concern is the health risks for teachers and school staff. A recent study with data for the UK found that, controlling for age, sex, underlying health conditions, and the number of adults in the household, primary school teachers did not have higher infection and mortality rates than the general population, suggesting that teachers are not at greater risk. The third concern is of the role of children and of schools in spreading the virus beyond the school premises. Existing evidence shows that the rate at which children transmit the virus, among themselves and to adults, is significantly lower than transmission rates among adults, so schools do not appear to be the super spreaders as was originally feared. Evidence from the United States, Spain, and Germany shows that the reopening of schools did not lead to an increase in the infection rate in these countries. Fourth, regarding the argument that children can spread the virus to older and more at-risk members of their household, a study including 12 million adults aged 65 or older living with and without children in the UK found no significant differences in infection rates and COVID-19-related hospitalizations and deaths between the two groups. All the evidence from high-income countries indicates that the health risks associated with opening schools are lower than what was perceived when the pandemic began, and also lower than the risks associated with opening restaurants, bars, markets, and other spaces that were (or are being) opened before schools in many countries.

During this past year, much information has also been generated about learning losses caused by the lack of face-to-face classes. Early theoretical models and simulations suggest significant losses, particularly among disadvantaged students. The World Bank estimates that in low- and middle-income countries, the percentage of children not able to read by age 10 (the Bank’s Learning Poverty indicator) will increase by 10 points, from 53% before the pandemic to 63%. Estimates using post-school closure standardized tests corroborate a negative impact on learning. Primary school students in Belgium, the UK, and the Netherlands show learning outcomes post-pandemic that are significantly lower than those in previous cohorts, and that this negative effect is highly concentrated among low-income students. The loss of learning happens after only a few months of school closures in a high-income context where most students have access to a device and to internet to follow online learning. There is no reason to believe that something magical happened in environments where closures have been longer (such as in Latin America or South Asia, see Figure 1) with no connectivity, and children spending at most a few hours a week following a class on TV. 


Figure 1: Number of school days by opening status in selected countries

Notes: the period covered goes from March 11, 2020 to February 2, 2021. The data source and opening status definitions corresponds to those of UNICEF.
Notes: the period covered goes from March 11, 2020 to February 2, 2021. The data source and opening status definitions corresponds to those of UNICEF.

The loss of learning even in rich country settings should not come as a surprise to those who have followed the debate on the effect of information technologies in education. Despite the great progress, there is no technology that can equate to teachers’ ability to generate learning among students. Information technology is only a complement, not a substitute, for the conventional teaching process – particularly among preschool and elementary school students. The importance of teachers, and the recognition of education as essentially a human interaction endeavor, is now even clearer. 

The pandemic has shown the importance of building resilient education systems that are prepared for the next pandemic, but also for school interruptions caused by natural disasters – such as earthquakes, floods, monsoons – or by conflict. Most students living in households with incomes under the poverty line in the developing world – roughly, the bottom 80% in low-income countries and the bottom 50% in middle-income countries – do not have the minimum conditions to learn at home. They do not have access to the internet, and, most of the time, their parents or guardians do not have the necessary schooling level or the time to assist them in their learning process. That digital divide must be closed to provide connectivity to poor households. But learning continuity also requires the presence of an adult, parent, guardian, or community worker assisting the student during the learning process while schools are closed. The rebuilding of the education systems will require investing in providing disadvantaged students the minimum conditions to learn at home.

False trade-offs?

Some might claim that the arguments above are invalid, even immoral, since, while it is true that learning is lost by keeping schools closed, what is to be gained are lives. However, education is also about saving lives, although in a less obvious way. The relationship between COVID-19 and a possible loss of life is evident, immediate, and very visible, while the lives lost due to school closures depend on indirect effects, often happening in the long-term. Schooling and learning are related to health, child marriage, early pregnancy, and life expectancy. Closing schools today reduces learning and educational trajectories, especially for the poorest, which reduces their future income, health, and life expectancy. Although the relationship between learning and life expectancy is not easily observed, and therefore difficult to internalize, opening schools also saves lives – the future lives of today’s children living in poverty.

What should we do?

The first step is to accept that, despite commendable efforts to implement a variety of multi-platform remote learning programs and initiatives by education authorities around the world, there is a learning loss caused by school closures. Some could interpret this as a failure of the remote learning schemes implemented throughout 2020. In some cases, they were not correctly implemented, but in many cases, they were the best possible response given the mediocre social, technological, and institutional conditions. Even before the pandemic, more than half of the students in developing countries were not achieving minimum proficiency. So, recognizing the huge learning losses does not imply a failure of the remote learning strategies implemented in 2020, but are, in fact, the result of many years of insufficient investment in education and insufficient attention paid to making the system more resilient. True, few people predicted a pandemic, and even fewer thought of establishing a system where learning processes are less susceptible to school disruptions. Consequently, the digital divide had not been closed, and schools, families, and communities were not prepared to handle remote activities efficiently. The discussion should now focus on how to minimize the impact of the pandemic on the learning of poor and vulnerable students who basically stopped their formal learning process in March of last year.

As sanitary conditions start improving, opening schools should be the priority, only below establishments of health services and food distribution, particularly those schools that serve poor and disadvantaged students. If they are open at least two or three days a week, even for a few hours, with a fraction of the students each day, social distancing, with adequate hygiene resources, with managed interaction with adults, the health risks can be largely contained, and learning can begin to recover.

When they open, schools will need a diagnostic test to identify the lag with which students are returning and implement a learning recovery plan. The evidence is overwhelming of the effectiveness of tutoring to improve learning among students who are lagging. The tutorials should be carried out in small groups, focused on foundational skills (math and language), and adjust the instruction to the level of competencies of each student – in other words, teach at the right level. This extra support to students cannot be done cheaply, therefore, countries must be realistic that more resources are needed. Unfortunately, about two-thirds of low-income countries are reducing their education budgets while many upper-middle and high-income countries are doing exactly the opposite. To this point, the UK recently approved an expansion in the education budget to finance, among other things, a national tutoring program, and the US is discussing the “Learning Recovery Act” to provide schools with additional resources to finance reengagement with at-risk students and help schools diagnose and close learning gaps. In addition, school calendars can be adapted to maximize learning time, and curriculums should be simplified to justify the focus on fewer subjects.     

By keeping schools closed while assuming that remote learning is a good substitute for face-to-face learning, we are simply making the poorest children pay for this with less learning today and less well-being tomorrow. What is clear is that learning recovery plans need to be developed as soon as possible with effective strategies and enough funding to support them. Failure to carry out the actions necessary to level out the playing field for the learning needs of poor and vulnerable children will only enable the costs of the pandemic to manifest, both in terms of poverty and inequality, for several generations to come.


Rafael de Hoyos

Program Leader for Human Development in the European Union, World Bank

Jaime Saavedra

Human Development Director for Latin America and the Caribbean at the World Bank

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000