COVID-19 lockdowns reduce children's probability to return to school: evidence from Nigeria

For many children worldwide, particularly those from low-income backgrounds, schooling represents the primary way to escape poverty and break the intergenerational transmission of poverty. However, in many low-income countries, aggregate income shocks often threaten children’s  education, which parents view as a costly investment. Moreover,  evidence shows that aggregate income shocks increase children’s vulnerability to child labor or child marriage (Corno and Voena (2016), Corno et al. (2020)) — two cultural practices known to undermine children’s schooling outcomes (Canagarajah and Coulombe, 1999; Field and Ambrus, 2008). When the COVID-19 pandemic erupted in early 2020, it was not surprising that concerns immediately arose over possible setbacks regarding progress achieved in educational outcomes (UNICEF, 2021b). Not only did the pandemic induce countries to enact protracted economy-wide lockdown measures, it also led to the temporary closure of schools. As lockdown measures became adverse shocks to household income and food security, there was a widespread concern that they might lead poor parents to pull children out of school permanently so that they could contribute to household income. However, since school closures were temporary, and in some locations children were able to maintain contact with school through distance learning (Dang et al. 2021), it is unclear whether such concerns were justified. 

In our study we use a unique dataset from Nigeria to test the hypothesis that COVID-19 school lockdown measures undermine children’s school resilience and reduce the number of students who returned to school after these measures were lifted.

To identify the effect that countrywide COVID-19 lockdown measures had on children's probability to attend school following schools' reopening, we compare a panel of school-age individuals observed just before schools' closure and just after schools' reopening and rely on the quasi-randomness of the occurrence of the pandemic. However, despite the plausible exogeneity of the COVID-19 shock, the potential simultaneous occurrence of other covariate shocks (e.g., climate shocks) can confound the identification of its effect on school attendance. Indeed, in Nigeria’s drought-prone western Sahel region, the concomitant occurrence of climate shocks such as droughts or floods is highly probable. Therefore, we control for exposure to climate shocks and shocks to household size both before the COVID-19 shock and post-shock to account for this potential challenge to identification.

We find that COVID-19 lockdown measures reduced children's school attendance across all of Nigeria. Our results show that the magnitude of adverse effect of lockdown measures increases with children's age. It is most prominent among children aged 15-18—those for whom schooling is no longer free and compulsory in Nigeria. This finding points to the fact that when hit with an adverse shock to their resources, families disproportionately stop sending children to school for whom education isn't compulsory, suggesting that such discontinuation is likely to lead to school dropout. We also find that among children aged 12-18 in the child-marriage prone North-West Nigeria, lockdown measures increase gender inequality by favoring boys to attend school, thus increasing girls' risk of becoming child brides.

We take the above results as suggestive evidence that in countries where cultural practices harmful to children are relatively common (e.g., child marriage and child labor), COVID-19 lockdown measures are likely to exacerbate children's vulnerability to these practices.  Of course, our data do not allow us to identify whether the adverse effect of lockdown measures on children's school resilience is permanent. However, as we find that this effect is more prominent among adolescents—a group of children whose economic value to parents becomes significant, there is ground for concern that it may indeed be permanent.