Monthly Archives: May 2023

Carte Blanche episode on SA’s Reading Crisis

This 13min Carte Blanche episode was aired on 30 April 2023.

10 Main Findings from PIRLS 2021 South Africa

On Tuesday the 16th of May 2023 the Progress in International Reading Literacy Study (PIRLS) released the findings from their 2021 survey. South Africa was one of 57 countries and regions that participated. is an independently administered, nationally representative assessment of reading comprehension among a sample of Grade 4 learners in South Africa. In 2021 the survey included 321 primary schools and tested 12,426 Grade 4 students between August to November 2021. South Africa has participated in PIRLS four times (2006, 2011, 2016 and 2021). The tests are set by the International Association for the Evaluation of Educational Achievement (IEA) and implemented in South Africa by the Center for Evaluation and Assessment (CEA) at the University of Pretoria. The tests are comparable over time and across countries, with all tests translated into the official languages of each country. In South Africa all 11 languages are tested. Children are tested in Grade 4 in whatever the language of instruction is from Grades 1-3 in the school (i.e. an isiZulu learner in an isiZulu school in KZN would write the test in isiZulu).  The SA 2021 Highlights Report is available here and the International 2021 Report is available here.

Here are the 10 main findings in my opinion (also in PDF here):

(1) In 2021 81% of Grade 4 learners cannot read for meaning in any language, up from 78% in 2016.

This means that only 19% of South African Grade 4 children could read for meaning in any language in 2021 (all 11 languages were assessed). Because PIRLS is a nationally-representative sample, of the 1,127,877 Grade 4 students in 2021, 914,000 could not read for meaning in any language. SA’s PIRLS score dropped from 320 (2016) to 288 (2021), approximately 0,8 years of learning.

(2) We have lost a decade of progress.

Between 2006 and 2016 the percentage of children that could not read declined from 87% (2006) to 82% (2011) to 78% (2016), but has now increased back to 81% (2021), wiping out a decade of slow progress and taking us back to 2011 levels of achievement.

(3) SA came last of all 57 countries, with the largest decline between 2016 & 2021.

South Africa had the lowest average Grade 4 test score of all participating countries. Of the 57 countries/regions, 33 had trend data from 2016 to 2021. Of those with trend data, South Africa experienced the largest decline in learning outcomes between 2016 and 2021.

(4) The average Gr4 child in SA in 2021 was 80% of a year behind their counterpart in 2016

SA’s PIRLS score dropped from 320 (2016) to 288 (2021). Given that approximately 40 points is equal to one year of learning, this is approximately 0,8 years of learning that we’ve lost on average.

(5) Northern rural provinces experienced the largest declines in reading.

Four provinces experienced declines of more than a full year of learning between 2016 and 2021. Given that 40 points amounts to one year of learning, this was as follows: North West (-2,4yrs), Free State (-1,6yrs), Mpumalanga (-1,2yrs) and Limpopo (-1yr). The three coastal provinces (WC, KZN and EC) experienced the smallest declines with WC showing the smallest decline (-0,4yrs).

(6) English and Afrikaans schools did not experience a decline between 2016 and 2021.

In comparison, most African language schools did decline, highlighting that the pandemic increased inequality between no-fee and fee-charging schools. (Note that the marginal increase in average scores for English and Afrikaans LOLT schools between 2016 and 2021 is not statistically significant. We do not yet have the standard errors for provincial averages since the DBE has not released the full South Africa PIRLS 2021 report (despite having access to it).

(7) Brazilian Grade 4’s are 3 years ahead of South African Grade 4’s.

The average score in Brazil was 419 points in 2021 compared to South Africa’s 288 points. The average Grade 4 child in South Africa is 3,3 years behind the average Brazilian Grade 4 child. In Brazil 61% of Grade 4’s could read at a basic level in 2021 compared to 19% in South Africa. Note that Brazil and South Africa have roughly the same GDP/capita ($7000/capita). See my Business Day op-ed on what we can learn from Brazil here.

(8) SA had the largest gender gap (pro-girl) of all 57 participating countries/regions, with SA girls 1,5 years ahead of SA boys.

The average Grade 4 girl in South Africa scored 57 points higher than the average Grade 4 boy, placing them about 1,5 years of learning ahead of their male counterparts. While girls outperform boys in all countries, the South African gap is more than twice the international average gap between boys and girls.

(9) SA does not currently have a credible or budgeted plan to catch up learning losses, despite experiencing the largest decline between 2016 and 2021.

It should be noted that PIRLS 2021 is the first nationally representative indication of learning losses to dateTo quote a recent research report reviewing DBE’s interventions relating to COVID-19 “There has been no attempt to recoup time in order to remediate learning losses, apart from very recent attempts in one province. The insistence on a largely business-as-usual approach to curriculum implementation fails to recognise and address the severe educational impact of the pandemic, especially on learners in the poorest communities (Hoadley, 2023: p.1). By contrast, many countries have multi-billion rand catch-up programs, for example Colombia’s PROMISE program (R3,5-billion), the Indian state of Gujarat’s GOAL initiative (R9,5-billion),  the Recovering Learning Losses program in northern Brazil (R4,8-billion), or the R7,3-billion ‘2023 Plan de Ractivación Educativa’ in Chile announced last month, acknowledging it will take at least four years to catch up the learning losses from COVID-19. Only one province (WC) has so far announced a budgeted plan for catching up learning losses (WCED’s R1,2bn ‘Back on Track’ program).

(10) A ‘generational catastrophe’

The new PIRLS 2021 reveal what can arguably be referred to as a ‘generational catastrophe.’ More than 4-million children in primary school have experienced more than half their schooling career in a disrupted state (either school closures or rotational timetables). Research on school closures from natural disasters like earthquakes in Pakistan and the Ebola crisis in West Africa all show that there are long term consequences to short term crises. These include lower educational attainment, lower earnings, higher unemployment and being more likely to be in lower skilled occupations in adulthood. This effect might even carry over to the children of the children affected by school closures, as happened with school closures in Argentina.

What can be done?

Evidence-based interventions to catch up learning losses. South Africa has the benefit of numerous interventions that have been proven to raise learning outcomes, even in no fee schools and in poorer provinces. These have been summarized in a recent book published by Oxford University Press (Spaull & Taylor, 2022)

(1)   Recruiting, training and equipping youth to be Teacher Assistants. Teaching assistants help teachers deal with large classes and different learning levels of children. A successful program in Limpopo selected youth using numeracy and literacy tests, trained them for four days face-to-face per term, equipped them with workbooks to use and monitored and supported them with TA-mentors (see Makaluza & Mpeta, 2022). This resulted in substantial improvements to literacy and numeracy equivalent to more than 1 additional year of learning. The government’s Presidential Youth Employment Initiative (PYEI) Basic Education Employment Initiative (BEEI) has established the organizational, political and financial feasibility of operating a teacher assistant programme at scale. The Limpopo programme demonstrates that it is also feasible to meaningfully impact learning outcomes and provides a model for recruitment, training and mentoring to achieve that. Subsequent iterations of the PYEI / Jobs Fund program could incorporate these learnings to catch up learning losses.

(2)    Rolling out anthologies of graded readers to all grade 1-3 children. Most children do not have basic texts needed to learn how to read in their home language at school or at home. This can be remedied using low cost anthologies of graded readers that cost R15/book/child.  For a brief period of 2 years (2019 and 2020) the Eastern Cape Department of Basic Education rolled out anthologies of graded readers – essentially a set of about 20 sequenced stories aimed at teaching children to read in their home language. The program was evaluated and shown to improve reading outcomes in isiXhosa for the children who received them compared to previous cohorts in the same schools who did not (see Ardington & Spaull, 2022).

(3)   Training teachers face-to-face and equipping them with comprehensive workbooks and teacher guides: (Ardington, 2023). Research in Limpopo shows that equipping learners with workbooks and teachers with teacher guides, in addition to four days of face-to-face training per term led to a 60% of a year of learning increase compared to business-as-usual schools.

(4)   Using teacher-coaches to support teachers on how to teach reading. The DBE’s Early Grade Reading Study (EGRS) shows that reading outcomes of Grade 1-3 learners improve after at least 2 years of an expert reading coach supporting and visiting teachers in the schools.


PDF of the above is available here.

My Business Day article today on PIRLS 2021: “Look to Brazil”

About 2,000 years ago Stoic philosopher Seneca was tutoring a Roman emperor and offered him the following advice: “If a man knows not to which port he sails, no wind is favourable.” The lesson here concerns the importance of planning and having a goal to know where you’re going. Unfortunately, when it comes to early grade reading in SA we are a ship at sea without a port in sight or plan in hand.  

On Tuesday we learnt that in 2021 81% of SA grade 4s couldn’t read a few simple paragraphs in their home language and answer basic literal questions. This is according to the Progress in International Reading Literacy Study (Pirls), a grade 4 test that was administered in all 11 languages in SA (SA highlights report here, international report here).

It’s not as if this was just a pandemic thing — the SA illiteracy rate in the last Pirls (2016) was 78%. This high rate of grade 4 illiteracy is unique to SA, at least among middle-income countries. The latest survey showed far lower rates in countries such as Egypt (55%), Jordan (53%) and Brazil (49%), all of which wrote the same test, at the same time, in their own languages.

Another shocking finding was that grade 4s in SA were three years behind grade 4s in Brazil, a country we like to compare ourselves to, perhaps optimistically. Brazil is an interesting case. Under former president Jair Bolsonaro the education budget was slashed, and he scrapped a successful national literacy programme.

When Lula da Silva was re-elected president earlier in 2023, his first move was to reverse the budget cuts in education and appoint Camilo Santana as education minister. Santana was the governor of Ceará, one of the poorest states in Brazil, but also the fastest-improving state on reading outcomes over the past 20 years. Ceará moved from poverty pariah to poster child, with countries now sending delegations to understand how they managed to get 84% of grade 3s reading for meaning.

Just two decades before, 40% of grade 3s in the state couldn’t read a single word. Going further, three months ago Lula created a new presidential secretariat dedicated specifically to books and literacy. For us in SA this is a parable of what’s possible with a president who follows words with actions, and who appoints ministers who are relentless about well-defined goals. 

You will recall that in 2019 President Cyril Ramaphosa announced in his state of the nation address that early grade reading was one of his administration’s “top five priorities” and that “within the next 10 years … every 10-year-old will be able to read for meaning”.

This was then codified in the medium-term strategic framework as one of “five fundamental goals”, alongside halving violent crime, increasing employment, eliminating hunger and promoting economic growth. But a goal is not a plan. Four years on from those lofty words, where are we on reading?

There is still no national reading plan, and if anything things have got worse. We have no plan to retrain our teachers, eliminate extreme class sizes, or resource classrooms with the specific books and tools teachers need to teach reading (graded readers, story books, alphabet friezes, etc). Instead, we have a patchwork document that saw the light of day only in May, and then only in response to Carte Blanche’s Promotion of Access to Information Act request.

We have no plan to retrain our teachers, eliminate extreme class sizes, or resource classrooms with the specific books and tools teachers need to teach reading.

This shambolic document is a laundry list of 47 activities. There are a few good ideas, but most are random fluff, including my favourite: “Partnerships with big business to include reading materials in food packages (eg ‘Did you know?’ messages in Chappies).” You can’t make this stuff up.

The plan lists targets for each year from 2019-2024, none of which has been reached. This Reworked National Reading Sector Implementation Plan 2019-2024 was released (under duress) only  in May 2023, a year before it will expire. 

A credible plan comes with credible milestones, a credible budget, a credible team and accountable reporting to political superiors. We have none of those things. Much of the department of basic education’s work on reading is outsourced to the National Education Collaboration Trust (NECT), a quasi-governmental project management office reporting to the director-general and minister. In part this is because the department itself is too slow or unable to execute the wishes of the minister.

The NECT has some good people working for it, but is under-capacitated and under-budgeted for the mandate it seems to have acquired. You cannot retrain 120,000 grade R to grade 3 teachers with a R37m budget, or even a R370m budget for that matter. To provide a sense of scale, we are spending R294bn on basic education in 2023, about 5% of GDP.

One province by itself, the Western Cape, is spending R400m in 2023 on Covid catch-up alone. The NECT, by contrast, is tasked with doing everything under the sun, including technology, maths, science, matric, reading, primary schools, high schools and Covid catch-up. 

Coming back to our stoic philosopher, it would seem as if the department has reframed the quote to be more palatable: “If a man knows not what harbour he seeks, any wind is favourable.” So, any activities, big or small, we’re here for you. Chappies partnerships, working with 150 schools here, 1,000 school libraries there, mobilising “Read to Lead” and encouraging teachers to “Drop all And Read”.

The problem with all this chatter is that it lacks prioritisation, it lacks accountability and it lacks budget. The presidency has been loath to hold ministers accountable for tangible, objectively verifiable progress towards a limited number of goals.  

We should heed the advice of history. That emperor was Nero, and he ignored Seneca’s advice. In fact, Nero forced him to commit suicide and went on to kill his own mother before burning the city of Rome. It’s where we get the expression “Fiddling while Rome burns”.

This feels especially apt when looking at the mess of 81% of grade 4s being unable to read for meaning, with no plan and no political will to get us out of it. And so, all I can say is: where’s the plan, minister Motshekga? And President Ramaphosa, are you a man of action like Lula, or does your word mean nothing?

This article first appeared in the Business Day on 18 May 2023.

Links I liked #34

  • Getting to SDG4: Useful 2023 UIS report on ‘Trends in learning proficiency in the last 20 years: How close are we to reliable regional and global SDG 4.1.1 trend statistics. Martin Gustafsson’s reports are always great and this is no exception. Sober, evidence-based and useful.
  • EGRA outcomes by gender: Nice new report (May 2023) by Fonseca et al out of RTI on “Girls have academic advantages and so do boys: A multi-country analysis of gender differences in early grade reading and mathematics outcomes.
  • The Latin-americanisation of SA schools: Last month Johnny Steinberg wrote an interesting op-ed for the Business Day ‘Bad education follows on the township exit of the middle classes. He makes the argument that because the middle classes have left public schools in South Africa, and that “public goods crumble when the middle classes exit”, SA no-fee schools will not improve. It’s an interesting argument and has been made before by Luis Crouch around the transition. I often use an extract of an email exchange I had with Luis in 2018 with my students to explain the nuances and trade-offs faced by the ANC around 1994 (see here). I was asking Luis to contextualise his use of the term “the threat of Latin-Americanisation of SA schools” and he explained it as “The notion that not having the middle classes vocally and publicly support public education as a matter of personal rather than abstract interest, because their own children were in private schools, and that this was deleterious to both accountability and budgetarily support, was fairly commonly held in Latin America at the time. Thus the “Latin-Americanization” notion, a term I may have coined in the South African context.” In most senses this did ‘work’ – SA has a relatively small private school sector. The DBE’s 2022 School Realities shows that only 5,5% of children in South Africa attend private schools (what we call ‘Independent Schools’). That figure is around 17% for most Low and Middle Income countries. Yet that hides the fact that fee-charging public schools are in a way semi-private (they allow parents to top-up public funds with private contributions). See Motala & Carel (2019) I don’t think Johnny’s argument is as clean or simple as he makes it out to be. For example, countries with similar levels of inequality – Brazil / Chile – have much better learning outcomes. It also doesn’t explain why, pre-pandemic, learning outcomes were actually improving quite rapidly (see Van der Berg & Gustafsson, 2019). Personally I think the political economy and bureaucracy dynamics (ANC-SADTU) explain more of the lock-jam than he gives credit to. But great to read thoughtful pieces in our media.
  • The Right to Read: The US Right to Read 2023 Trailer – Glad to see campaigns like this getting momentum. Our own Right to Read & Write published by the South African Human Rights Commission (SAHRC) makes similar arguments – essentially that learning to read is a fundamental human right that all are entitled to by virtue of their birth and common humanity.
  • Chile’s COVID-19 catch-up program: Tutorías 2023 with a great promo video:
  • WCED’s Back on Track Program: On Thursday the WCED launched it’s R1,2-billion ‘Back on Track‘ program covering tutoring, Saturday classes, holiday camps etc. There is some detail in the presentation but we’ll have to wait for the full roll-out plan with all the details. Very encouraging to see a costed and budgeted plan for COVID catch up that matches the scale of the devastation.
  • Private schools & PPPs in LMIC: Crawfurd, Hares and Todd (2023) published their paper “The Impact of Private Schools, School Chains and PPPs in Developing Countries” – they conclude that: “The search resulted in over 100 studies on low-cost private schools and PPPs, with a large majority being on low-cost private schools. Our meta-analysis shows moderately strong effects from private schooling, although the limited number of experimental studies find much smaller effects than quasi-experimental studies. This advantage, though, is not nearly enough to help most children reach important learning goals. Turning to policy goals, we find that the private school advantage has not translated to public private partnerships, which have shown limited value in improving quality. They can however represent a lower-cost means of increasing access to school. We also find that private school chains perform little better than individual private schools and have little scope for achieving meaningful scale.”
  • Tracking FLN donor spending: SEEK Development launched their “Donor Tracker for Foundational Literacy and Numeracy (FLN)” The dominance of the United States when it comes to bilateral funding for FLN is significant. Of the $505million spent on FLN in 2020, 68% ($344mil) came from the United States (see below). If you do a deep-dive into US funding specifically you see that the total funding for FLN declined sharply between 2019 ($434mil) and 2020 ($344mil), a 21% decline i na single year. FLN funding is also now a smaller proportion of total funding in 2020 (27% of total) compared to 2019 (30% of total). These are worrying trends since it suggests that the biggest bilateral donor is shifting away from FLN. Great tool to look at funding trends over time.
  • RISE’s Education Systems Course: The folks at Rise have created an Open Access “Education Systems Course” with lectures and videos. Useful syllabus to catch up on readings in this area.
  • School feeding in SA: Given the recent furore over the DBE school feeding scheme (NSNP) failures in the Eastern Cape and KwaZulu-Natal, I was reminded of a 2016 report I came across evaluating the NSNP in two districts in the Eastern Cape. Frustrating that even “stock standard” decades long interventions like the NSNP can go pear shaped without proper monitoring.
  • Racial dynamics in SA: Justine Burns and co-authors published a paper in the AER on “Interaction, Stereotypes, and Performance: Evidence from South Africa.” Fascinating paper: “We exploit a policy designed to randomly allocate roommates in a large South African university to investigate whether interracial interaction affects stereotypes, attitudes and performance. Using implicit association tests, we find that living with a roommate of a different race reduces White students’ negative stereotypes towards Black students and increases interracial friendships. Interaction also affects academic outcomes: Black students improve their GPA, pass more exams and have lower dropout rates. This effect is not driven by roommate’s ability.”
  • A Moral Education: Garth Greenwell writes an essay for the Yale Review that is at times piercing and other times rambling on “In praise of filth.” I liked the initial discussion about the modern turn (or return) to moralising art and the problems with woke culture categorising art “Within the small world of people who care about literature and art, the culture is as moralistic as it has ever been in my lifetime: witness our polemics about who has the right to what subject matter, our conviction that art has a duty to right representational wrongs, that poems or novels or films can be guilty of a violence that seems ever less metaphorical against an audience construed as ever more vulner­able. We have a sense that the most important questions we can ask about a work of art are whether and to what extent it furthers extra-artistic aims, to what extent it serves a world outside itself. The idea that artists should make what they feel compelled to make, regardless of such considerations, that in fact art should be pro­tected from responsibilities of this kind, seems part and parcel of a discredited Romantic model of the artist as exempt from workaday morality, licensed by genius to act badly, or at least to disregard the claims of others.”
  • Learning losses and resisting “back to normal” policies: The New York Times published an article “Parents Don’t Understand How Far Behind Their Kids Are in School” (via Brahm Fleisch). Their conclusion is especially apt in South Africa where the modus operandi has been very much “back to normal” in most provinces: “As enticing as it might be to get back to normal, doing so will just leave in place the devastating increase in inequality caused by the pandemic. In many communities, students lost months of learning time. Justice demands that we replace it. We must find creative ways to add new learning opportunities in the summer, after school, on weekends or during a 13th year of school. If we fail to replace what our children lost, we — not the coronavirus — will be responsible for the most inequitable and longest-lasting legacy of the pandemic. But if we succeed, that broader and more responsive system of learning can be our gift to America’s schoolchildren.”