Scout Briefing: 9/10/26
Today’s strongest development is the release of PISA 2025—the Programme for International Student Assessment—which gives a new national and international view of 15-year-old reading performance. It adds weight to a pattern Scout has been tracking: early-literacy reform matters, but the U.S. reading problem clearly extends into adolescence, and the challenge is broader than foundational decoding alone.
1. U.S. 15-year-old reading performance declined again
Finding. PISA 2025 shows U.S. reading performance lower than in 2022 and below 2018 levels, placing it among the lowest U.S. results since participation began in 2000. Compared with 2015, the share of U.S. students performing below the minimum reading proficiency level increased by seven percentage points.
Why it matters. This is fresh evidence that the national literacy problem persists well beyond K–3. Improvements in beginning-reading policy have not yet translated into a broad improvement in adolescent reading.
2. The U.S. simultaneously has many strong readers and a substantial weak-reader population
Finding. About 74% of U.S. 15-year-olds reached at least PISA Level 2 in reading, compared with 69% across the Organisation for Economic Co-operation and Development (OECD). Thirteen percent reached the highest reading levels, more than double the OECD average of 6%. Yet roughly one-quarter remained below Level 2.
Why it matters. Averages hide an important market and instructional reality: the U.S. is capable of producing highly proficient readers while still leaving a large group unable to perform basic reading tasks reliably. That points toward differentiated intervention rather than one explanation for national performance.
3. The new U.S. PISA results need an important methodological caution
Finding. OECD flagged the United States because student response rates, school participation, and exclusion rates did not fully meet PISA technical standards. OECD says resulting bias cannot be ruled out or confidently estimated in either direction. More than 40% of participating U.S. students also did not receive the background questionnaire.
Why it matters. The broad decline deserves attention, but individual percentages and comparisons should not be treated as more precise than the data allow. This is a useful reminder not to turn one assessment release into a simple national narrative.
4. The international decline makes the U.S. problem part of something larger
Finding. Across OECD countries, PISA reading and mathematics performance fell to the lowest averages recorded by the assessment. One in five 15-year-olds now performs poorly across reading, mathematics, and science, up from 16% in 2022.
Why it matters. Weak U.S. reading results cannot be explained solely by U.S.-specific Science of Reading policy or curriculum choices. Pandemic disruption, attendance, technology use, reading habits, socioeconomic conditions, and broader changes in adolescent learning all remain plausible contributors.
Contrary evidence. This complicates any argument that stalled U.S. performance proves recent structured-literacy reforms have failed. Many of those reforms target younger students who have not yet reached PISA age.
5. PISA adds evidence to the technology debate—but not a simple anti-screen conclusion
Finding. OECD reports that students tended to perform less well when digital or artificial-intelligence use became excessive or when devices were used for leisure during school. More than one in four students internationally said classmates were distracted by devices in most or every science lesson. At the same time, OECD says students need both strong academic foundations and the ability to solve problems using digital tools.
Why it matters. The finding is not “technology is bad.” It is closer to purpose, dosage, and context matter. Digital competence and core academic competence appear complementary rather than interchangeable.
CRSL connection. This is consistent with—but does not prove—the distinction CRSL is making between teacher-support technology and unnecessary student-facing screen use.
6. Federal literacy policy is becoming more concentrated around evidence standards
Finding. On September 9, the U.S. Department of Education summarized a major shift in current grantmaking: nearly $500 million in recent Education Innovation and Research grants, $256 million in continuing Comprehensive Literacy State Development awards, upcoming teacher-preparation and educator-development awards, and new literacy grants are being tied explicitly to evidence-based literacy priorities.
Why it matters. Whatever happens politically to federal education structures, current funding rules are reinforcing a market in which programs increasingly need to describe not just instructional philosophy but the quality of evidence supporting their components.
7. The federal definition of evidence-based literacy is notably broader than phonics
Finding. Federal Education Innovation and Research criteria define evidence-based literacy as explicit, systematic instruction that includes phonological awareness, phonics/decoding, oral language, vocabulary, language structure, fluency, comprehension, writing, and knowledge-rich materials. Eligible evidence must generally come from well-designed experimental or quasi-experimental studies.
Why it matters. This is an important corrective to shorthand descriptions of federal Science of Reading policy. At least in formal grant language, the definition extends across the reading system rather than reducing reading instruction to foundational skills.
CRSL connection. This resembles CRSL’s broader system architecture more closely than a phonics-only definition does, though CRSL’s individual evidence claims still have to stand on their own.
8. Teacher-support technology is now an explicit federal research target
Finding. The Institute of Education Sciences is currently funding development of HOL Advisor, a K–2 platform designed specifically to help teachers and school leaders interpret student data, build instructional plans, access training, track implementation, and receive instructional recommendations. The project is still in development/testing, not established evidence of effectiveness.
Why it matters. This is useful behavioral evidence that “teacher-facing technology” is not merely a local design idea. Federal research dollars are testing technology whose primary user is the educator rather than the student.
Caution. One development grant is an early signal, not evidence that this model improves outcomes.
9. Federal policy is also pushing states to build their own evidence systems
Finding. The Department of Education released a new Learning Agenda Playbook on September 9 after piloting it with Arkansas, Georgia, Idaho, Kansas, New Hampshire, Tennessee, Utah, and Wisconsin. The model asks state agencies to identify priority problems, formulate questions, build evidence, and use it to redirect policy and resources.
Why it matters. If states take more control over education policy, vendors may face more state-specific evidence expectations, not fewer. State approval processes, evaluation frameworks, and implementation requirements could become increasingly important parts of market access.
10. Multilingual literacy remains an unresolved edge of the Science of Reading movement
Finding. Two 2026 research syntheses argue that much of the Science of Reading evidence base was developed around monolingual English readers and that multilingual learners may require explicit attention to cross-language transfer, home-language knowledge, vocabulary, and metalinguistic strengths in addition to foundational English reading instruction.
Why it matters. This is not contrary evidence against explicit phonics or structured literacy. It is evidence against assuming that practices validated primarily with monolingual populations automatically constitute a complete instructional model for multilingual students.
Scout is noticing…
Observation — adolescent literacy can no longer be treated as a secondary extension of the early-reading story. PISA adds another independent national signal that large numbers of students reach adolescence without adequate reading proficiency. The policy and intervention market beyond K–3 therefore deserves attention on its own terms.
Observation — “evidence-based literacy” is becoming both broader and more demanding. Federal definitions increasingly include comprehension, oral language, vocabulary, writing, and knowledge alongside decoding, while funding structures are asking for stronger research evidence and implementation evaluation.
Observation — states may become more—not less—important to literacy vendors. Federal policy is deliberately building state evidence capacity while directing substantial literacy funding through state agencies. The likely result is a more fragmented approval and implementation environment across the country.
Early signal — teacher-facing technology is beginning to appear as a recognizable development category. The federal HOL Advisor project, recent district limits on student AI, and PISA’s caution about excessive digital use are converging around a possible distinction between technology that helps educators teach and technology that asks students to spend more instructional time with a device. The evidence is not yet strong enough to treat this as a settled movement.
Assumption Watch
“Technology can strengthen instruction when it serves the teacher rather than displacing the teacher.” — External evidence continues to make this plausible, but outcomes remain unproven. Current evidence increasingly supports asking where the technology sits and what human work it supports: OECD cautions against excessive student digital use while still valuing digital competence, and federal research is testing teacher-facing tools designed to reduce instructional and implementation friction. None of this yet proves that teacher-facing technology produces better student outcomes; that remains the useful test.
