South Africa's silent dropout crisis: what happens between Grade 1 and matric

Research · June 2026 · 22 min read

Every January, South Africa celebrates its matric results. In 2025, the country celebrated a record-breaking 87.3% pass rate for the class of 2024 — the highest in the nation's history. Ministers posed for photographs with top achievers. Headlines praised the hard work of teachers and learners. The celebration was real, and deserved.

But there is another number that rarely appears in those headlines. Of the 1.22 million learners who started Grade 1 in 2013, only 614,562 passed matric in 2024. That is a real pass rate of 50.25% — and it is the number that tells the full story of South Africa's education system.

Two numbers, two different stories

The official pass rate is calculated from the pool of learners who actually wrote the final exams. It excludes the hundreds of thousands who never reached that point — those who dropped out in Grade 8, repeated Grade 10 twice and left, or simply disappeared from the system in the years between.

The Build One South Africa party's acting spokesperson Roger Solomons called the real figure a "flashing red light" for the country's education system. Equal Education, an advocacy group, noted that of 1,222,851 learners who started Grade 1 in 2013, only 724,156 even enrolled as full-time Grade 12 candidates in 2024 — meaning over 498,000 learners did not even make it to write the exam.

"Cracks in the education system disrupt learning, affect performance and push learners to disengage, putting them at high risk of dropping out before matric."

Where the system loses learners

Research from the Zero Dropout Campaign identifies several pressure points where the system tends to break down. The transition from primary to high school — the Grade 7 to Grade 8 shift — is a well-documented danger zone. Many learners who were performing adequately in primary school find themselves suddenly overwhelmed in secondary school, often without the support structures needed to catch up.

A practice sometimes called "gatekeeping" compounds the problem. Schools under pressure to maintain pass rates will sometimes hold learners back — not to help them, but to protect the school's statistics. Repeated retention is demoralising. Many learners simply stop coming.

Financial pressure is another factor. In South Africa's poorest quintile schools, learners often face hunger, long commutes on foot, and family circumstances that require them to earn income rather than attend school. When the choice is between feeding younger siblings and writing a mathematics test, the classroom loses.

Before Grade 1: where the pipeline actually starts

Every explanation so far assumes the crisis begins somewhere inside the school system — Grade 7, Grade 10, the moment a learner is held back once too often. After the 2025 results were released, Minister Gwarube pushed back on that framing directly. The deepest crisis in South African education, she argued, is not visible in the matric results at all. It sits in early childhood development — years before any child sets foot in a Grade 1 classroom.

The data behind that argument comes from the Thrive by Five Index 2024, the most comprehensive national assessment of early childhood development South Africa has produced, covering more than 5,000 four-year-olds across all nine provinces. Among children enrolled in early learning programmes, only 42% are developmentally "on track" across early learning, physical growth, and social-emotional functioning. A further 28% are falling behind, and 30% are falling far behind. For the roughly 30% of four-year-olds not enrolled in any early learning programme at all, only 18% are on track — meaning the least-served four-year-olds in the country arrive at school already structurally behind, before a single lesson has been taught.

What happens to a child who arrives in Grade 1 already behind on early numeracy and fine motor skills — the two domains where the Thrive by Five Index found the weakest outcomes, with only 33% and 29% of enrolled children on track respectively? In South Africa, the system's default response is repetition. Research drawing on national survey data finds that by the time learners reach Grades 10 to 12, 52% have repeated a grade at least once, and 9% of Grade 12 learners have repeated a grade three or more times. The same research notes that academic gains from retention tend to fade within a few years — a "washout effect" that leaves the repeating learner older than their peers, no further ahead academically, and statistically more likely to leave school before matric.

There is a recent complication worth noting. A December 2025 Stellenbosch University analysis found that repetition in the very early grades — Grade 1 and Grade 4 — does appear to carry lasting benefit, with repeaters retaining a measurable marks advantage years later. The picture that emerges is one where repetition might genuinely help a five- or nine-year-old catch up, but by the time the same mechanism is applied to a 16- or 17-year-old in the FET phase, it functions less as remediation and more as a holding pattern before the learner leaves entirely. Grade repetition currently absorbs roughly 8% of South Africa's entire basic education budget — close to R20 billion a year — spent disproportionately on the secondary phase, where the evidence for its benefit is weakest.

None of this is a separate crisis from the one described above. It is the same crisis, observed earlier. The learner who disengages in Grade 8 after being "gatekept" into repeating Grade 7 is, in a meaningful number of cases, the same child who arrived in Grade 1 already among the 58% who were not developmentally on track at age four. The dropout crisis does not begin when a teenager stops coming to school. Statistically, it begins in an early learning programme — or the absence of one — years before that teenager's name ever appears on a school register.

The throughput rate is getting worse

What makes the 2024 data particularly alarming is the decline in retention. The throughput rate — the proportion of learners who actually complete the 12-year schooling cycle — dropped to 64.5% for the class of 2024. That is a regression from previous cohorts, not an improvement. The country is passing more of the learners who stay, but fewer learners are staying.

"While graduation rates may be improving, the overall retention and completion of secondary education remain significant challenges."

— Zero Dropout Campaign, January 2025

Even for those who stay: the gateway subject collapse

Throughput measures whether a learner reaches matric at all. It says nothing about what that learner is equipped to do once they arrive. The 2025 results contained a second crisis, quieter than the dropout numbers but arguably just as consequential for the economy: a collapse in both participation and performance in the subjects that determine access to STEM careers and university — what the Department itself calls "gateway subjects."

Only 34% of the 2025 matric cohort wrote Mathematics at all, down from 38% the year before; the rest took Mathematical Literacy, a parallel subject that does not carry the same weight for STEM admission. Of the minority who did write Mathematics, the pass rate fell from 69% in 2024 to 64% in 2025 — a near five-percentage-point decline in a year the overall pass rate reached a record 88%. And in 2025, 464 public schools did not offer Mathematics as a subject at all.

Researchers attribute this to the same upstream forces described throughout this piece. Analysts reviewing the 2025 results pointed to weak foundations accumulated from the early grades as the primary driver — the same foundations the Thrive by Five Index measures at age four, and the same foundations grade repetition is meant to repair but rarely does. Some researchers describe a "pass-rate hostage" dynamic: schools under pressure to protect their headline pass rate have an incentive to steer weaker learners away from Mathematics and into Mathematical Literacy before Grade 10, when subject choices are finalised — improving the school's statistics while narrowing, permanently, that learner's access to STEM study.

This is dropout of a different kind. The learner is still enrolled. They will likely pass matric — South Africa's overall pass rate has never been higher. But the door to engineering, medicine, computer science, and the skilled trades the economy says it needs has been quietly closed, years before that learner ever sits the exam that confirms it. Of the roughly 778,000 learners who reached the 2025 matric cohort, well under 10% achieved a strong pass in Mathematics. The headline pass rate and the pipeline of STEM-ready graduates are, increasingly, two different numbers measuring two different things.

What this means for iSkool

iSkool was not built to replace teachers, or to solve poverty, or to paper over infrastructure crises. None of those problems are solved by software. What iSkool can do — what it was specifically designed to do — is reduce the friction that causes learners who are still in school to quietly disengage.

When a learner misses an assignment because they did not know it had been set. When a teacher cannot track who has fallen behind in a class of 55 learners. When parents have no visibility into what is happening at school. These are the moments where small tools, applied carefully, can make a meaningful difference at the margin — and in education, the margin is where hundreds of thousands of learners live.

The dropout crisis is systemic. iSkool addresses one thread of it. But one thread, pulled consistently, matters.

The gender angle: girls disappear first

The dropout numbers are bad. For girls, they are worse. South Africa's own Department of Basic Education reported that approximately 150,000 girls between the ages of 10 and 19 became pregnant in the 2022/23 school year alone — and one in three of those girls did not return to school afterward. Only 52.5% of girls in South Africa complete their secondary education, compared to higher rates for boys.

Pregnancy is the most visible cause, but it rarely stands alone. The Commission for Gender Equality's 2024 study on adolescent school dropout — conducted across KwaZulu-Natal, the Eastern Cape, and Limpopo — found that the factors compounding pregnancy into permanent dropout include stigma from teachers and peers, lack of healthcare access, and parental opposition to girls returning to class after giving birth. The DBE's own policy allows pregnant learners to remain in school, but policy and practice diverge sharply in under-resourced communities.

Gender-based violence adds another layer. The DBE has described GBV as a "second pandemic" within schools, with between 20% and 34% of children experiencing contact violence before the age of 18. In April 2022 to March 2023 alone, over 42,000 rape cases were reported to the South African Police Service. For girls who experience sexual violence — whether at school, en route to school, or at home — attendance becomes a calculated risk, not an entitlement.

Then there are the responsibilities that never appear in any official statistic. The Zero Dropout Campaign's analysis of the 2023 General Household Survey found that about 7.2% of children under 18 are not attending school due to pregnancy, marriage, or caregiving duties — looking after younger siblings, an ill parent, or elderly grandparents. Of that group, only 0.1% are boys. The rest are girls, pulled out of class not by choice but by the logic of households where caring work defaults to the daughter.

Stats SA's own Q1 2026 data captures the downstream effect. The NEET rate for young women aged 15–24 is 39.2% — 3.2 percentage points higher than for young men, and rising. Women without a matric certificate face an unemployment rate of 42.8%, compared to 37.0% for men with the same educational profile. The school-level dropout is not the end of the story; it is the beginning of a labour market story that compounds year by year.

The other half of the gender story: the boy child

The gap that opens at age four does not stay at age four. The Thrive by Five Index found that 48% of girls were developmentally on track at four years old, compared with 37% of boys — an eleven-point gap before either child has started school. By the time the same cohort reaches matric, that gap has not closed. It has reversed direction and grown.

In the Class of 2025, girls made up 56% of matric candidates; boys made up just 44%. That is not a story about girls succeeding at the expense of boys — it is a story about two different failure modes operating on the same age group. Pregnancy used to be the dominant reason girls left school permanently; policy changes that allow pregnant learners to remain enrolled, and to return after giving birth, have measurably improved girls' retention. Boys face no equivalent re-entry pathway. Once a boy leaves the system between Grade 10 and Grade 12 — usually after repeating a grade, and usually after a period of disengagement rather than a single triggering event — he is far less likely than a girl in the same position to ever come back.

Minister Gwarube has used blunt language to describe the consequence: the boy child, she has said, "is being left behind." The Department has linked this specifically to grade repetition, loss of motivation, and — in the starkest framing the Minister has used publicly — vulnerability to recruitment into gangs once a boy is no longer in a classroom for most of the day. Where the girls' story is, at its core, a story about reproductive health policy and caregiving expectations colliding with school attendance, the boys' story is a story about what happens to a teenage boy's time, structure, and sense of belonging once the system has told him — through repetition, through being steered out of gateway subjects, through simply falling behind — that he no longer fits.

Both halves of this story point to the same conclusion from a different angle. A system that loses a majority of one gender to early caregiving and reproductive health barriers, and a substantial share of the other gender to repetition-driven disengagement and informal exclusion, is not failing along a single axis. It is failing in at least two distinct ways that happen to land, almost evenly, on either side of a single statistic — 590,000 young people who did not reach the exam hall.

The language wall: what happens at Grade 4

South Africa has eleven official languages. Most of its learners speak none of the two that dominate its school system. English is the home language of roughly 9% of South Africans. Afrikaans is spoken by around 13%. Yet English or Afrikaans is the language of instruction in the vast majority of the country's classrooms from Grade 4 onward — which means that by the time most learners reach the first years of secondary school, they have been sitting in lessons conducted in their second or third language for half their schooling life.

The policy is not accidental. South Africa's Language in Education Policy of 1997 allows children to be taught in their home language in Grades R through 3, then switches the language of learning and teaching to English or Afrikaans from Grade 4. In practice, more than 80% of learners from African language backgrounds make this switch at Grade 4 — entering a classroom where the language of instruction, the textbooks, the assessments, and eventually the matric exams are all in a language they learned as a subject, not as a mother tongue.

The 2021 PIRLS results — South Africa's most recent international reading assessment — show the consequences of that wall. Of the 57 countries assessed, South Africa recorded the largest decline in reading outcomes. Eighty-one percent of Grade 4 learners cannot read for meaning in any language. The 2030 Reading Panel, convened by former Deputy President Phumzile Mlambo-Ngcuka, found that at current trajectories it will take South Africa 80 years for all children to read for meaning — and that when the system comes under fiscal pressure, the learners hardest hit are those in no-fee schools, already managing the language switch without the private tutoring or well-resourced classrooms that make it survivable for others.

"Multilingual education is vital; learners who receive instruction in their mother tongue during their early years perform better when they move over to English from Grade Four. It not only improves academic outcomes but also fosters inclusivity and respect for cultural diversity."

What makes this crisis particularly invisible is that it sits beneath the literacy numbers, not next to them. When a Grade 8 learner cannot engage with a history lesson or understand a mathematics word problem, the system records the outcome — a poor mark, a repeated grade, an eventual dropout — but rarely records the cause: that the learner was asked to master abstract content in a language they never fully acquired. The 2030 Reading Panel estimates that illiteracy costs South Africa's economy approximately R119 billion annually. It does not estimate what that figure would look like if the language of instruction were subtracted from the equation.

Two systems, one department: the township vs. Model C divide

Open a Department of Basic Education annual report and you will find one education system. Drive fifteen minutes in any South African city and you will find two.

The informal shorthand of "Model C schools" refers to the formerly whites-only government schools that, after 1994, were allowed to charge school fees and govern themselves through school governing bodies. They remain technically public schools under the DBE. They receive the same national curriculum, employ teachers on the same pay scales, and write the same matric exams. But they operate in a fundamentally different financial reality — one in which parents can contribute R30,000 to R80,000 per learner per year in fees, fund their own computer labs and sports facilities, and hire additional staff when the department's allocation falls short. Research published in a 2024 study on machine learning and matric outcomes found that 64.1% of schools achieving 100% matric pass rates belonged to Quintile 5. Among schools with pass rates below 20%, 56.3% were Quintile 1.

Township schools — predominantly Quintile 1 to 3, designated no-fee schools — receive more government funding per learner than their wealthier counterparts, precisely because they cannot charge fees. The national allocation for a no-fee school learner stands at roughly R1,242 per year. A well-resourced Model C school collecting R30,000 per learner from its parents receives a total of more than 24 times that amount, before counting supplementary fundraising, donor contributions, or the kind of parental time and involvement that is simply not available to a family whose adults work two jobs and commute three hours daily.

The irony is structural. The quintile system was designed to direct more government money toward the poorest schools. It does this. But it cannot close the gap that fee income creates on the other side. A 2020 study on school governing bodies in Gauteng found that more than 530 Quintile 4 and 5 schools in Gauteng granted fee exemptions worth R631.6 million in a single financial year — yet received only R13 million in compensation from the province. The Model C system is simultaneously subsidising poor families who can access it and extracting resources from the public pool that cannot be replenished.

The Gauteng Department of Education is now contending with a R444 million budget shortfall for 2026, on top of a R160 million shortfall forecast for the following budget period. In January 2026, reports emerged that Quintile 5 schools faced funding realignments of up to 64% — a figure the department disputed but which nonetheless prompted panic among governing bodies that have built their staffing models on fee income supplemented by state allocations. The two systems are not merely unequal. They are becoming structurally incompatible.

What actually happens to the 590,000

South Africa's matric results tell the story of those who stayed. The 590,000 young people who never reached the exam hall have a different story — one that unfolds not in percentage points but in the quarterly labour force data, the NEET statistics, and in the communities where the consequences of school failure become intergenerational.

The most recent data is unambiguous. In Q1 2026, Statistics South Africa reported that 60.9% of young people aged 15–24 were unemployed — the highest rate of any age group. Among youth without a matric certificate, the unemployment rate stands at 51.6%. Even completing matric provides limited protection: youth with matric face a 47.6% unemployment rate. The credential that the country celebrates every January is, for many who hold it, not a key to the labour market but merely a slightly lower floor.

The NEET rate — young people not in employment, education, or training — reached 37.6% for youth aged 15–24 in Q1 2026, and 45.6% for the broader 15–34 age group, both rising year on year. The Southern Africa Labour and Development Research Unit (SALDRU) found that more than half of unemployed youth in 2025 had been searching for work for a year or longer. Long-term structural unemployment has become the default condition for a generation, not a temporary state between opportunities.

"Many young people are not NEET because they reject education, but because the system rejects them. These bottlenecks create a compounding effect. Students who miss out on placement do not vanish — they wait, re-apply, attempt informal pathways, and in worse cases become discouraged and drop out of institutional engagement entirely."

For those who do find work without a matric, the options are narrow. The Trade sector — retail, wholesale, hospitality — absorbs the largest share of employed youth, followed by community services. These are not pathways to financial stability; they are pathways to the bottom of a wage structure where the national minimum wage is the ceiling, not the floor. Stats SA's 2024 labour market dynamics report found that about 90% of South Africa's 8.2 million unemployed persons had either matric or had not completed secondary education — meaning the labour market's floor has effectively risen to the level of tertiary qualification, leaving an enormous mass of young people with nowhere to stand.

The downstream consequences extend beyond individual earnings. SALDRU's analysis notes that the NEET rate has increased steadily from 29% in 2015 to 34% by 2025 — a decade of accumulating exclusion, across cohort after cohort, in which each year's dropout class joins the one before it. The UCT Children's Institute has identified the two central long-term risks of this accumulation: the perpetuation of intergenerational poverty, and the implication of a large population of young people disconnected from both the economy and the civic institutions that give it structure. Those risks are not abstract. They are already visible in the unemployment figures, in the crime statistics, and in the political temperature of a country where the 15-to-34 age group makes up 49.7% of the working-age population.

The policy moment this all sits inside

Everything described in this piece — the early years gap, the repetition cycle, the gendered dropout patterns, the gateway subject collapse — is now playing out against the backdrop of the most significant change to basic education law in a generation. The Basic Education Laws Amendment (BELA) Act was signed by President Ramaphosa in September 2024, and after a brief delay over its most contested provisions, was proclaimed in full in December 2024. It is now, through 2025 and 2026, working its way into how schools actually operate.

Three of BELA's changes connect directly to the story told here. First, Grade R — the reception year before Grade 1 — is now a compulsory part of basic education for the first time, formally extending the schooling system one year earlier and into the exact developmental window the Thrive by Five Index measures. Second, and most directly relevant to dropout, BELA strengthens attendance enforcement: parents and guardians are now legally obligated to ensure a child attends school until Grade 9 or age 15, whichever comes later, with non-compliance carrying penalties that include fines or imprisonment of up to twelve months. For the first time, a learner quietly disappearing from the register is not only a school-level statistic — it is, on paper, a compliance matter with legal consequences for the adults responsible for that child.

Third, BELA introduces new financial and governance accountability requirements for School Governing Bodies — the same bodies that, as described earlier, manage the fee income that separates a Quintile 5 school's resources from a no-fee school's R1,242-per-learner allocation. None of these changes solves the underlying problems this piece has described. Compulsory Grade R does not, by itself, fix the fact that only 42% of four-year-olds are developmentally on track. Attendance enforcement does not feed a hungry learner or undo a language barrier that began at Grade 4. But for an education system that has spent years measuring its own dropout crisis without a clear legal mechanism for addressing it, BELA represents the first attempt in decades to make attendance — the precondition for everything else in this story — a matter of law rather than aspiration.