Global student performance hits new low as PISA reveals sharp decline in reading, math skills

Approximately 29% of OECD 15-year-olds lack foundational skills in reading, math, and science needed for basic life tasks, limiting future educational and economic opportunities.
Nearly 30 percent of teenagers lack foundational skills for basic life.
The 2025 PISA study found that almost a third of 15-year-olds in wealthy nations cannot reliably read, calculate, or interpret data.
Mark

So this PISA test—it's been running since 2000. Has performance always been declining, or is this a recent crash?

Mimi

There's been a downward trend in many countries for years, but this latest drop is particularly steep. Reading scores fell 28 points in OECD countries since 2015 alone. That's about a year's worth of learning, gone.

Luke

But the study itself says there's no direct link between school closures and reading decline. So we can't just blame COVID. What else is actually documented as a cause?

Mimi

The study points to digital distractions, declining interest in reading books, and now AI use in schoolwork. More than 40 percent of students use AI to draft writing assignments regularly, and those students score lower in science.

Mark

Germany specifically—why is it falling so hard?

Mimi

Germany's been sliding since 2000, but the recent acceleration is striking. Between 2022 and 2025, reading dropped 15 points and math fell 11. They're now 24th in reading globally.

Luke

Is that because German schools are worse, or because the test measures something different now? Are we comparing apples to apples across decades?

Mimi

The test methodology is designed to be consistent. But you're right to push on that. The study does note that Germany has extreme variation between schools—differences account for half of all performance variation. That's a structural inequality problem.

Mark

So some German schools are excellent and others are struggling?

Mimi

Exactly. The study calls it academic segmentation. Schools reserved for higher achievers, and then separate institutions for lower-achieving students. When that sorting overlaps with family background, it amplifies inequality.

Luke

The study says parents asking kids about school once a week correlates with a 26-point science gain. That's a nice finding, but it's correlation, not causation. Are we suggesting parental curiosity is the solution to a systemic crisis?

Mimi

No. The study frames it as one thing parents can do. But it's acknowledging that the real challenge is structural—education systems aren't adapting fast enough to technological and social change.

Mark

What comes next? Are countries actually going to change their systems?

Mimi

That's the open question. The study lays out the problem clearly. Whether policymakers respond with real reform is up to them.

  • Nearly 30 percent of fifteen-year-olds in OECD countries cannot reliably read a simple text, interpret a basic graph, or reason through an everyday transaction — a generation underprepared for the world it is about to enter.
  • Reading scores across OECD nations have fallen 28 points since 2015 — the equivalent of an entire year of learning simply gone — with Iceland and Slovenia losing more than 50 points alone.
  • Germany, still carrying the wound of its first PISA shock in 2000, has continued its slow descent, now ranking 24th in reading and 26th in mathematics, with stark inequality between schools suggesting the system is fracturing along social lines.
  • Over 40 percent of students report using AI chatbots to write assignments weekly, and those students score lower in science — raising urgent questions about whether technology is supplementing learning or quietly replacing it.
  • Policymakers face pressure to move beyond damage control: the study points to structural failures that predate COVID-19, and only five countries — Qatar, Brunei, Cambodia, the Philippines, and Zambia — managed to improve at all.

Each generation entrusts its children to schools with the hope that knowledge will become capability — that learning will translate into a life well navigated. The 2025 PISA study, measuring the real-world skills of 760,000 fifteen-year-olds across 91 countries, suggests that trust is eroding: reading, mathematics, and science scores have fallen sharply across wealthy nations, with nearly a third of young people now lacking the foundational skills needed for everyday life. Germany, long haunted by its first PISA reckoning in 2000, has slid further still, while the broader data reveals not merely a pandemic disruption but a structural unraveling that digital distraction, inequality, and institutional inertia have quietly deepened over years.

On September 8, the 2025 PISA study delivered a stark verdict on global education: fifteen-year-olds are performing worse than they have in decades. Testing more than 760,000 students across 91 countries, the assessment measures not rote memorization but practical capability — whether a teenager can extract meaning from a text, interpret a graph, or reason through a basic transaction. By that measure, the world's schools are falling short.

Across OECD nations, reading scores have dropped 28 points since 2015 — roughly one full year of lost learning. Mathematics fell 22 points, science seven. Iceland and Slovenia saw reading collapse by more than 50 points. Only five countries improved: Qatar, Brunei Darussalam, Cambodia, the Philippines, and Zambia — all still below the OECD average, but moving in the right direction. Nearly 30 percent of OECD fifteen-year-olds are now classified as low performers across all three subjects, a cohort whose path to further education and economic participation is narrowing.

Germany's trajectory is particularly painful. The country's first PISA results in 2000 were a national shock; the years since have brought only incremental recovery and renewed decline. Between 2022 and 2025 alone, German reading scores fell 15 points and mathematics 11. The country now sits 24th in reading, 26th in mathematics, and 23rd in science. More troubling than the rankings is what lies beneath them: in Germany and nine other nations, differences between schools account for more than half of all variation in student achievement — a sign of a system sorted into tiers, where family background and school quality compound each other.

The causes resist simple explanation. School closures during the pandemic played a role, but the decline predates COVID-19 and shows no clean correlation with how long schools remained shut. Digital distraction, waning interest in reading, and now artificial intelligence have all entered the picture. More than 40 percent of students reported using AI chatbots to draft assignments at least weekly — and those students scored lower in science than peers who did not. The study stops short of drawing firm conclusions but flags the pattern as a concern.

Amid the weight of the data, the study offers one quietly hopeful finding: parents who ask their children once a week or more about what they learned at school see a 26-point gain in science performance, even after accounting for household wealth. Curiosity and presence matter. But the scale of what the numbers reveal demands more than parental attentiveness — it demands that governments reckon honestly with what their schools are, and are not, preparing young people to do.

On September 8, the world's most comprehensive measure of student learning delivered a sobering verdict: fifteen-year-olds across the globe are performing worse than they have in decades. The 2025 PISA study, which tested more than 760,000 students in 91 countries and economies, found sharp declines in reading, mathematics, and science—a downward trajectory that has alarmed education officials and policymakers from Tokyo to Berlin.

The Program for International Student Assessment, known as PISA, does not measure how well students have memorized textbook facts. Instead, it asks whether they can actually use what they know to solve real problems. Can a teenager read a simple text and extract information from it? Can she understand the relationship between price and quantity well enough to shop for groceries? Can he interpret a graph or make sense of basic scientific data? These are the skills the test probes. Since its debut in 2000, PISA has become the global standard by which nations judge their educational health.

The numbers this year paint a picture of systemic erosion. Among the 38 wealthy nations in the OECD, reading scores have fallen 28 points since 2015—a decline equivalent to roughly one year of learning. Mathematics has dropped 22 points on average. Science, the focus of this year's assessment, fell seven points. The damage is not evenly distributed. Iceland and Slovenia saw reading scores plummet by more than 50 points. Only five countries bucked the trend: Qatar, Brunei Darussalam, Cambodia, the Philippines, and Zambia all posted gains, though their absolute performance remains below the OECD average. The study acknowledged their progress as noteworthy given the broader collapse.

The human consequence is stark. Nearly 30 percent of fifteen-year-olds in OECD countries are now classified as low performers across all three subjects tested. These are not students who score poorly on an advanced exam. These are teenagers who cannot reliably extract meaning from straightforward text, who struggle to interpret basic graphs, who lack the mathematical reasoning needed for everyday transactions. For millions of young people, the foundation for further education and economic participation is crumbling.

Germany's experience is particularly acute. When PISA first ran in 2000, Germany's results shocked the nation—its students scored below the international average, a humbling discovery for a country with deep educational traditions. The country has been sliding ever since. In the 2022 test, German fifteen-year-olds hit a new low. This year, they fell further still. Between 2022 and 2025 alone, reading scores dropped 15 points and mathematics fell 11 points. Germany now ranks 24th globally in reading, 26th in mathematics, and 23rd in science—barely holding ground above the OECD average. The top positions in all three subjects are dominated by Asian countries. The United States ranks 13th in reading and 27th in mathematics, just below Germany.

Behind these national averages lies a more troubling pattern. In countries like Iceland, Cambodia, and Finland, student performance is relatively consistent from school to school—less than 10 percent variation overall. But in Germany and nine other nations, differences between schools account for at least half of all variation in student achievement. This suggests a system divided into tiers: schools for high achievers and schools for everyone else. When such academic sorting compounds with family background and economic circumstance, the study warns, it amplifies inequality. Excellence concentrates in pockets while the broader system fails to ensure consistent learning across the whole population.

Why is this happening? The data alone cannot explain the decline, but the study's authors point to patterns that predate the pandemic. School closures during COVID-19 may have contributed, but there is no direct correlation between how long schools remained closed and how badly reading scores fell. The deterioration reflects deeper structural problems. Digital distractions pull students away from sustained reading. Interest in books has waned. And now artificial intelligence has entered the picture: more than 40 percent of students reported using AI chatbots to draft writing assignments at least once or twice a week, and those who did so scored lower in science on average than their peers who did not. The study concludes that education systems are struggling to adapt to rapid social, technological, and economic change.

For policymakers, the challenge is not simply to recover lost ground but to reimagine what schools must do in a transformed world. The study offers one modest suggestion for parents: asking children once a week or more about what they learned at school correlates with a 26-point gain in science performance, even after accounting for family wealth. Small gestures of curiosity and support matter. But the scale of the problem demands more. The question now is whether governments will act with the urgency the data demands.

Where academic segmentation compounds with social background, it can amplify inequalities, producing pockets of excellence while failing to ensure consistent learning outcomes across the education system as a whole.
— PISA 2025 study authors
The challenge for policymakers is not simply recovering lost ground, but to ensure schools meet the rapidly changing demands of the world.
— PISA 2025 study summary
Contattaci Domande frequenti