Mauritian pupils are learning less. The world is too, but some countries found a way back
School Certificate results are the lowest in at least ten years, and Mauritius's first PISA test since 2010 shows more than half of 15-year-olds below the basic level in maths and reading. Rich countries are slipping too. A few turned it around, and the evidence on what works is clearer than the debate suggests.
Photo: Karsten Ratzke, CC0
In short
The national exams are slipping. 69.3% of School Certificate candidates passed in 2025, the lowest in at least ten years, and fewer than half got a credit in maths. Only about 1 child in 3 who starts primary school reaches the Higher School Certificate.(7, 8, 13)
The international test agrees. In PISA 2025, a test of 15-year-olds in 91 countries and economies, 44% of Mauritian pupils reached the basic level in maths and 48% in reading, against 50% and 53% in 2010. Science improved. Mauritius is still the best of the five African countries tested.(1, 2)
The rich world is falling too: the OECD average is the lowest PISA has ever recorded, and the slide began before COVID. Devices used for fun at school and less reading for pleasure go with lower scores.(1)
Money alone is not the answer. Spending on education has fallen from 3.5% to about 3% of GDP, and trials show the best value comes from teaching children at their level and giving teachers step-by-step lesson plans.(16, 20)
It matters for jobs: half of Mauritius's unemployed have no School Certificate, and economists find that what pupils actually learn, not how long they stay in school, drives a country's growth.(14, 19)
What the exams say
Every year, Mauritian pupils sit four big national exams. The PSAC, or Primary School Achievement Certificate, closes primary school at the end of Grade 6, around age 11. The NCE, or National Certificate of Education, comes at the end of Grade 9. The School Certificate, the SC, is the Cambridge O level taken at the end of Grade 11, around 16. The Higher School Certificate, the HSC, the Cambridge A level, closes secondary school at the end of Grade 13 and opens the door to university.
The School Certificate tells the clearest story. Between 2015 and 2019 the pass rate drifted down from 72.5% to 70.9%.(5, 16) During the pandemic years it jumped, to 85.7% in 2020-21 and 78.5% in 2022, then fell back: 73.7% in 2023, 72.7% in 2024 and 69.3% in 2025, when 8,130 of 11,730 candidates passed.(6, 7, 16) That is the lowest pass rate of any year since at least 2015. The jump was not a real gain in learning, and the fall since is a return to a slow decline that was already under way.
The detail is worse than the headline. A credit is the grade above a simple pass, and it is what sixth forms and employers look for. In maths, 48.5% of candidates got a credit in 2017; in 2025 it was 42.4%. In French it fell from 50.9% to 48.0%, and in English it rose to 63.8% in 2021 before falling back to 54.1%.(8) In other words, fewer than half of 16-year-olds leave Grade 11 with a good mark in maths.
The HSC followed the same path: 75.0% passed in 2019, 92.3% in 2022, then 84.4%, 78.9% and 79.0% in 2025.(4, 6, 16) Fewer young people are sitting it at all: 5,794 candidates in 2025, against more than 7,500 in 2024 and more than 10,000 in the early 2010s.(9) Many pupils also drop out on the way: when the Director of Audit followed the 16,967 children who started Grade 1 in 2012, only 5,857, about 1 in 3, reached the HSC.(13)
Lower down, the picture is flat rather than falling. In the PSAC of December 2025, 76.9% of pupils passed at the first sitting, about the same as the year before; boys did worse than girls, 73.5% against 80.4%, and in schools in the Priority Education Zones, the poorest areas, only 45% passed.(10) At the NCE in Grade 9, 70.4% passed in 2025, below the 75% the ministry was aiming for.(11, 13) And the Extended Programme, the four-year track for pupils who fail the PSAC, has failed them again: in 2025 only 78 of its 1,460 pupils passed the NCE, 5.3%.(12) A new Foundation Programme in literacy and numeracy replaced it in 2025, and 134 of 173 secondary schools offered it in March 2026.(4)
What the world test says
National exams can change their rules from year to year. An international test lets a country compare itself over time and with others. PISA, the Programme for International Student Assessment, is run by the OECD, the Organisation for Economic Co-operation and Development, a club of 38 mostly rich countries. Every few years it tests 15-year-olds in maths, reading and science, asking them to use what they know on real problems: reading a timetable, judging a claim, working out a discount. Scores are set so that the rich-country average was 500 when each subject was first tested. About 20 points is roughly one year of schooling.(1)
Mauritius took part for only the second time. In 2025, 5,360 pupils from 128 schools sat the test, standing for about 11,700 fifteen-year-olds; the first time was in 2010.(2) The results, published on 8 September 2026, are mixed.(1, 2) Science rose from 417 to 438 points, about a year of schooling more. Reading is unchanged at 405. Maths fell from 420 to 407, more than half a year less. The OECD average is 482 in science, 461 in reading and 463 in maths, so Mauritian 15-year-olds are 56 points behind the rich-country average in maths and in reading, and 44 in science. The OECD says about 20 points is roughly a year of learning, but warns that this is only a rough guide when comparing countries.(1, 2)
The share of pupils at the basic level, which PISA calls Level 2, matters more than the average. In maths it means, for example, converting a price into another currency; in reading, finding the main idea of a medium-length text. 44% of Mauritian pupils reached it in maths, down from 50% in 2010, and 48% in reading, down from 53%, and 60% in science, up from 53%. The OECD says none of these three changes is big enough to be sure it is real; the rise in the average science score is.(2, 3) The OECD averages are 65%, 69% and 74%.(1) 35.1% of Mauritian pupils were below the basic level in all three subjects at once, against 19.7% across the OECD.(1) Almost nobody reached the top levels: 1% in maths and 3% in science.(2)
Two points put this in perspective. Mauritius is the best of the five African countries tested: Morocco scored 358 in science, and Zambia, Kenya and Rwanda scored between 317 and 337.(2) It ranks 47th of 90 countries and economies in science, in the OECD's own table.(2) And part of the narrowing gap with the rich countries is not Mauritius improving but the rich countries falling, the ministry's own press release notes.(2)
The test also asked pupils about school life, and some answers stand out.(2, 3) 29% of Mauritian pupils said they were bullied a few times a month or more, against 20% in the OECD. 29% had skipped a school day in the two weeks before the test (22% in the OECD), and 57% had arrived late. Pupils from better-off homes scored 94 points more in science than poorer pupils, a gap similar to the OECD average of 85. Girls were 20 points ahead of boys in reading. On the positive side, Mauritian pupils are more curious than the OECD average and more of them say they keep going when work gets hard: 75%, against 60%.
Why the rich world is slipping too
Mauritius is not alone. PISA 2025 found the lowest OECD averages it has ever recorded in all three subjects. Since 2015 the average has fallen 22 points in maths, just over a year of learning, and 28 points in reading, about a year and a half.(1) Even the strongest systems fell between 2022 and 2025: Singapore lost 12 points in maths, Japan and Finland lost ground in every subject, and Hong Kong fell by 19 to 25 points.(1) Across the OECD, 20% of pupils are now below the basic level in all three subjects, up from 16% in 2022.(1)
COVID made it worse, but it is not the main cause: the OECD says the weakening began before the pandemic in many countries, with reading peaking around 2012.(1) The report points to several changes that happened at the same time as the fall, while stressing that it cannot prove which one caused it. Fewer teenagers read for pleasure, and more of their reading is short texts on screens; the share of 'hasty readers', who skim and answer too fast, almost doubled between 2018 and 2025.(1) Using phones and tablets for fun during school goes with substantially lower scores, while moderate use for learning goes with better ones.(1) Pupils who use AI chatbots to summarise texts for them score lower in science.(1)
Phone bans are the popular answer. About half of OECD pupils now attend schools that ban phones, up from about a third in 2022, and pupils there report a little less distraction. But PISA found no link between countries adding bans and countries improving their scores.(1) Mauritius already bans phones in schools attended by 79% of pupils, far above the OECD's 49%, and its pupils still spend about one and a half hours a day on devices at school for leisure, more than the OECD average.(2, 3) A ban on paper is not the same as attention in class.
Not everyone fell. Türkiye has gained about 60 points in science in ten years, and the United Kingdom improved in science since 2022.(1) Beijing, Shanghai, Jiangsu and Zhejiang in China, Singapore, Japan, Korea and Estonia stay at the top.(1) Estonia, a country of 1.4 million people, is the closest thing to a model for a small state: it scores 508 in maths, far above the OECD average.(1) Wealth helps, but only up to a point: once a country spends about 85,000 US dollars on each pupil between ages 6 and 15, more money stops going with higher scores. Luxembourg spends about 288,500 dollars per pupil and scores below Ireland, France and Italy in science, though they spend less than half as much.(1)
How to get the standards back: five lessons from economics
Economists from different schools agree on more than the public debate suggests. Here is what each lens adds.
1. Growth economics: count what pupils learn, not years in school. Two economists, at Stanford University and the University of Munich, compared countries over 1960 to 2000. Test scores one standard deviation higher, roughly the gap between an average and a strong school system, went with about 2 percentage points more growth in income per person every year. Once skills are counted, the number of years spent in school barely matters.(19) They estimate that getting every child into lower secondary school with at least PISA Level 1, one step below the basic level, would make the economy of an upper-middle-income country, the World Bank's group for Mauritius, 16% bigger on average over the next 80 years.(19) The World Bank's 2026 human capital measure gives Mauritius 201 points out of 325, against 259 for Poland, and says years of schooling, learning outcomes and university completion explain most of the gap.(18)
2. Development economics: teach children at their level. In 2023 a panel of experts backed by the World Bank, UNICEF, USAID and the UK reviewed more than 13,000 studies to find the best value in education. One of its 'great buys' is Teaching at the Right Level: grouping pupils by what they can actually do, not by their age, for part of the day, and teaching the basics until they catch up.(20) Structured pedagogy, giving teachers step-by-step lesson plans and matching books, gave Kenyan pupils more than a full year of extra reading progress.(20) The panel's 'bad buy' is buying computers and tablets on their own.(20) Mauritius's Extended Programme, where 95 in 100 pupils failed the NCE, is a warning: separating weak pupils without changing how they are taught does not work.(12)
3. Keep at it for years: the Mississippi lesson. Mississippi, one of the poorest US states, saw its Grade 4 reading scores rise until they reached the US national average in 2019.(21) Its 2013 reading law, which from 2015 makes pupils repeat Grade 3 if they still cannot read, often gets the credit. But an analysis by Mississippi First, an education group in the state, finds that most of the rise came between 2005 and 2019, and that the law can only explain the gains from 2017 on, when most of the progress had already been made; it links the rise to a long run of reforms, including No Child Left Behind, new state standards and a broad reform drive since 2012.(21) The lesson is steady work on reading over many years, not one law. The Budget 2026-27 includes a literacy and numeracy programme for Grades 1 and 2, the right place to start.(17)
4. Behavioural economics: small nudges on attendance. Nearly 3 in 10 Mauritian pupils had skipped a day in the previous two weeks.(2) Trials in the United States found that sending parents simple text messages about missed classes and missed homework raised attendance and cut failed courses, at a very low cost.(24) In one trial, pupils attended about 12% more classes and failed 27% fewer courses. Parents often think their child's grades are better than they are and do not see how much homework is skipped; telling them the real picture changes behaviour.
5. Public finance: spend smarter, and earlier. Public spending on education fell from 3.5% of GDP in 2015-16 to about 3.0% planned for 2023-24.(5, 16) Education and training takes 8.9% of the government's budget in 2025-26, up from 8.0% the year before but below 11.6% in 2015-16.(4, 5) The Director of Audit found Rs 17.4 billion invested over five years in pre-primary, primary and secondary schools.(13) The Budget 2026-27 adds Rs 2.6 billion for the two education ministries, Rs 824 million for buildings and equipment and Rs 10 million for an anti-bullying campaign; the Government Teachers' Union says it does nothing to fill staffing gaps in primary schools.(17) The evidence points to spending early, in the first school years, where the gaps open.(20)
There is also a choice to make now. In December 2025 the government put two ways of admitting pupils to secondary school out for consultation: choosing pupils for Grade 7 on their PSAC and abolishing the NCE, or keeping the NCE and admitting pupils to national schools at Grade 10. Under both, every child gets a place in a regional school.(23) The lessons above suggest the bigger question is not which exam sorts pupils, but how many of them reach the basic level before any exam at all.
Why it matters for jobs and the economy
School results show up later in who struggles to find work. At the end of 2025, half of Mauritius's 31,800 unemployed had no School Certificate, and 3,000 had not even passed the PSAC.(14) In the second quarter of 2026, 19.3% of young people aged 16 to 24 who wanted a job could not find one, more than three times the overall rate of 5.5%.(15) An economy that wants to move into finance, technology and high-end services needs people who read, calculate and solve problems well. As MoBriefs explained on 10 November, the jobs AI changes first are routine office jobs; the ones that remain need stronger skills, not weaker.
The global picture shows the cost of doing nothing. Before COVID, 57% of 10-year-olds in low- and middle-income countries could not read and understand a simple text; afterwards the World Bank and UNICEF estimated 70%, and 89% in sub-Saharan Africa, with a cost of about 21 trillion US dollars in lost lifetime earnings worldwide.(22) Mauritius is far better placed than most of the continent. But it is not where its income says it should be: the World Bank scores it just below the average of countries with similar incomes (201 points against 204),(18) and its 15-year-olds are scoring lower in maths than in 2010.
What it means for you
Take Kevin, 11, in a primary school in a Priority Education Zone. On the 2025 figures, fewer than half of pupils in schools like his pass the PSAC at the first try.(10) If he fails, he will spend secondary school in a track where, until now, 95 in 100 never passed the next national exam.(12) And across his whole year group, only about 1 in 3 children who start Grade 1 reach the HSC.(13) Nothing in that path is decided by talent alone. It depends on whether someone notices at 7 or 8 that he cannot yet read well, and teaches him at his level.
For parents, the research points to what helps at home: reading for pleasure goes with better results, and using screens for fun during school hours goes with worse ones, even if PISA cannot prove which causes which.(1) Knowing how often your child is absent or late matters: in Mauritius almost 3 in 10 pupils had skipped a day in the previous two weeks.(2)
The one thing to do: ask your child's school how it checks whether children in Grades 1 and 2 can read and count, and read with your child for 15 minutes a day, with phones out of the room.
