Every three years, the OECD tests hundreds of thousands of 15-year-olds around the world in reading, math, and science, and every three years the results make headlines for a news cycle. This time the headline is earned. PISA 2025, released September 9, recorded the lowest OECD-average performance ever measured in all three subjects — with reading taking the hardest hit, down roughly 14 points since 2022 on top of 2022's own record decline. Average reading performance across the OECD now sits around 461, down from roughly 500 when the test began in 2000. That's a decline of about 7–8% over 25 years, and most of it happened in the last decade. If you're raising or teaching a child who struggles to read, this global story and your kitchen-table story are connected — but not in the way the loudest takes suggest.
What exactly declined, and when?
Three things are worth separating. First, the size: reading fell about 14 points across the OECD since 2022 and math about 9 — large movements by the standards of an assessment where a school year of learning is often estimated at around 20 points. Second, the shape: scores were roughly flat from 2000 to about 2012, then began sliding, with the steepest drops after 2018. This was not a slow 25-year leak; it's a recent, accelerating one. Third, the spread: the decline is sharpest across OECD member countries, while a handful of systems — Singapore, Japan, Korea among them — have held up far better, and science was broadly stable among non-OECD participants. The slide is real and broad, but it is not uniform, which matters when we get to causes.
Didn't the US move up the rankings?
Yes — and that's the least comforting good news in the report. The US scored 490 in reading and 490 in science, both comfortably above the OECD averages of 461 and 482, and 463 in math, exactly at the OECD average. But US reading also declined from 2022, and US math sits near its lowest level ever; the country climbed the table mainly because others fell faster. There's a further caveat the celebratory posts leave out: the OECD attached a data-quality warning to the US results because school response rates were incomplete, saying it isn't possible to rule out bias — only Albania received a similar flag. So the honest US summary is: relatively strong position, weakening absolute performance, measured with less certainty than usual. Our NAEP analysis tells a consistent domestic story: American 13-year-olds are reading at levels last seen decades ago.
Are phones and screens the cause?
Here's where careful reading matters. The timing genuinely lines up — the decline begins around 2012 and accelerates after 2015, tracking smartphone ubiquity — and the OECD itself points to digital distraction and shrinking attention spans as contributing factors. Survey data also consistently shows teens reading less for enjoyment, with 2026 coverage tying the slump directly to surging screen time. But the OECD frames the screen link as an association, not proven causation, and the steepest recent drop (2018–2022) is heavily confounded by COVID school closures. Anyone declaring the case closed is going beyond the evidence — in both directions. What we can say with confidence: sustained reading of longer texts has declined, something is eroding it, and screens are the leading suspect with means, motive, and opportunity. We take a practical position on what that means for young readers in our companion piece on hands-on reading foundations.
What does a global average mean for a child who struggles?
Averages hide tails. When an average falls this far, the students who were already behind — including many with dyslexia — are the ones with the least margin, and the systems serving them get noisier: more struggling readers overall means screening lists grow, intervention slots tighten, and a child whose difficulty has a specific, identifiable cause can disappear into a general malaise ("everyone's scores are down"). That's the risk to guard against. Dyslexia is not a screen-time problem — it's a brain-based difference in processing written language that responds to explicit, structured teaching — and a falling tide must not become a new reason to wait and see. If your child is struggling, the playbook is unchanged and unambiguous: know the signs, request an evaluation in writing (our toolkit), and insist on evidence-based intervention.
What can families and teachers actually do with this news?
- Use it, don't fear it. PISA headlines open doors: school boards and legislators pay attention for a season. It's a good month to ask your district what it's doing about structured literacy — bring the State of Dyslexia report for your state.
- Rebuild daily reading at home. The most defensible response to the reading slump requires no legislation: books in hands, every day, on paper where possible. Our read-together guide has the routine.
- Keep long-form reading alive. The skill PISA measures — sustained comprehension of extended text — is exactly what short-burst scrolling doesn't exercise. Protect stretches of unhurried reading, by eye or by ear.
- Don't let averages excuse individuals. If a teacher or administrator waves off your child's struggle as part of a general trend, the answer is: general trends don't get evaluations — children do, and the law provides for it.
Frequently asked questions
What is PISA and who takes it? The Programme for International Student Assessment is an OECD test given every three years to representative samples of 15-year-olds in dozens of countries, measuring reading, math, and science. It tests applied skills — like understanding and evaluating extended texts — rather than curriculum recall, which is why it's watched as a gauge of deep literacy.
Did smartphones cause the global reading decline? Not proven. The timing correlates — the decline begins as smartphones spread — and the OECD cites digital distraction as a contributing factor, but it frames the link as association, and COVID disruption confounds the recent drops. Reduced time spent reading long texts is well documented; the exact causal chain is not settled.
My child's scores are fine — should I care about PISA? Mostly as context and leverage. PISA moves policy conversations, and the practices it argues for — evidence-based reading instruction, daily reading, protected long-form attention — benefit every child, and matter most for children with dyslexia, who have the least room to absorb weak instruction.
Source: "PISA 2025: students' reading and mathematics performance declined sharply across the OECD" — OECD press release, Sep 2026 · "U.S. ranking on PISA improves as other countries decline" — Chalkbeat, Sep 8, 2026 · "Teen reading slumps to worst this century amid screen-time surge" — U.S. News, Sep 8, 2026