Counting the Cost of Lost Schooling in South Africa

• In Event: Educational Impact of COVID-19: A Comparative Perspective Examining Finland, Mexico, Norway, Portugal, Singapore, and South Africa

Abstract

The purpose of this analysis is to estimate the loss of contact school time in 2020, the disruptions for opportunities to learn and to speculate the effects on achievement scores.
South Africa is a complex society with high levels of poverty and income inequality overlain with deep fractures of race, gender and class. The young education system is structurally fragile as a result of the injustices of apartheid. Education is key for development. We started in 1994 with very low achievement levels, and over 25 years have made gains and our education system is now described as ‘low and unequal achievement’ (Reddy et al,2020).
In contexts of high household poverty, parents and society view schools and classrooms as the institutions to equalise opportunities for students from poorer homes. Through face-to-face quality educational engagements, students are offered an opportunity to learn. Covid disrupted the opportunities to learn for the majority of the school going population. The Covid related school closures led to disrupted OTL with the effect of lower achievement scores, i.e. learning losses. These losses are higher for the poor and the more vulnerable.

Given the unprecedented nature of the pandemic, the methodological uncertainties to estimate learning losses, the limited empirical data we have we will review other studies that predicted learning losses by relying on ‘suggestive extrapolations’ from previous studies. These learning losses are predicted in terms of loss of a share of a year of schooling, reduction in learning gains expressed as percentage loss, loss of future earnings or loss of years of learning. Maldonado and De Witte (2020) calculated the effects of school closures on standardized scores in Belgium using panel test data and administrative data which measured students’ socio-economic status. We wanted to measure the extent of learning losses in South Africa.
In the absence of primary data to estimate the extent of learning losses for South Africa we posed the question: If the TIMSS 2019 achievement instrument was administered to grade 9 learners in October 2020 (call it TIMSS 2020) what would their mathematics achievement score be?

South Africa participated in TIMSS 2019 and collected nationally representative data. We also collected data on home assets (to construct a home SES indicator) and data on the fee-paying status of the school (proxy for the school SES). We will draw on Maldonado and De Witte (ibid) and other methodologies to develop an “informed speculation” model to estimate the effects of disrupted opportunities on TIMSS achievements in 2020.
Using this model, we will present 2020 national achievement estimates as well as achievement levels for resourced and under-resourced schools. This estimate will provide an indication of the extent of learning losses in the South African educational system.
Since 1994, the South African educational achievements have slowly improved. This study contributes to measuring the achievement effects (and hence learning losses) of the coronavirus on a fragile, low-performing and unequal educational system. This will contribute to understanding effects locally in order to build back, as well as how South African learning losses resonates with that in other countries.

Author

• Vijay Reddy, Human Sciences Research Council, South Africa