America’s school mood just hit a 27-year low
Only 32% of U.S. adults say they are satisfied with K-12 education. Yet 66% of parents are satisfied with their own child’s education. Fresh PISA and NAEP data show why both views can coexist—and why the national argument is more complicated than a single grade.
September 10, 2026 · United States
American public opinion about schools has reached a striking contradiction. The national system is being graded harshly, but the school a family actually knows—the teachers, classrooms, routines and report cards attached to a real child—still receives a much better score. That gap is not a footnote. It is the central fact of the education story in September 2026.
Gallup reported this week that 32% of Americans are satisfied with the quality of K-12 education in the United States, down 11 percentage points in two years and the lowest level in the organization’s trend dating to 1999. Sixty-seven percent are dissatisfied. The historical comparison is stark: satisfaction averaged 44% over the full period, reached 53% in 2004 and was at 51% as recently as 2019.
At almost the same moment, the OECD released its 2025 Programme for International Student Assessment, better known as PISA. The U.S. results give critics real material: reading performance among American 15-year-olds fell from 2022 and, together with mathematics, remains below 2018 levels and among the lowest the United States has recorded in PISA. But the international comparison also resists a collapse narrative. U.S. students still scored above the OECD average in reading and science and close to the OECD average in mathematics.
Meanwhile, the newest long-term NAEP results—the federal assessment often called the Nation’s Report Card—show younger students recovering ground that older students have not. Nine-year-olds improved in both reading and math from 2022 to 2025; their reading performance was no longer significantly different from the pre-pandemic 2020 level. Thirteen-year-olds, by contrast, showed no significant improvement in reading or math from the prior assessment and remained below pre-pandemic performance.
Put together, the polling and testing data describe neither a system that is simply failing everywhere nor one that can dismiss the public’s frustration as perception. They describe a country with uneven recovery, persistent reading weakness, political distrust, anxiety about preparation for adulthood and a surprisingly durable reservoir of trust at the schoolhouse level.
The most important number may be the 34-point gap
National mood versus local experienceGallup’s most revealing comparison is 32% versus 66%. Parents of children currently in kindergarten through 12th grade are about twice as likely to say they are satisfied with their own child’s education as they are to say the same about U.S. K-12 education overall. The 66% local satisfaction figure is itself a record low in Gallup’s series—so there is no basis for pretending families are unconcerned. But it remains dramatically higher than the national rating.
This “local-national gap” appears in other areas of American life: people often view institutions in the aggregate more negatively than the specific professionals or services they encounter. Education magnifies the pattern because a national judgment is formed through test-score headlines, partisan conflict, school-safety debates, social media, curriculum disputes and stories from districts thousands of miles away. A parent’s judgment of a particular school is built from different evidence: whether a child wants to go to class, whether assignments make sense, whether teachers answer emails, whether reading is improving, whether the school feels safe and whether the family sees a plausible path to graduation.
The gap matters because it changes the diagnosis. If almost every parent believed the school directly in front of them was broken, the policy problem would look like immediate institutional failure. Instead, many Americans appear to believe that the national system is in trouble while their own school remains functional or even good. That can reflect genuine geographic variation. It can also reflect the distance between national political narratives and local experience. Most likely, both forces are operating at once.
National confidence is weak
The 32% satisfaction reading is not ordinary dissatisfaction. It is the lowest Gallup has measured in 27 years and is paired with a record 67% dissatisfaction rate.
Local trust has not disappeared
Two-thirds of K-12 parents still say they are satisfied with their own child’s education, and 68% rate their school positively for adapting curriculum, schedules and instruction for the future.
The political pattern reinforces the point. Gallup found satisfaction at 35% among Democrats, 32% among independents and 25% among Republicans. Since 2024, satisfaction fell 19 points among Democrats, 10 points among independents and six points among Republicans. That does not mean the decline is merely partisan; independents and Republicans also moved lower. But it does mean any account of the national mood that ignores politics is incomplete.
How to read the Gallup figures
The national satisfaction figures come from a Walton Family Foundation/Gallup telephone poll conducted Aug. 3–24, 2026. A separate probability-based online survey conducted Aug. 3–17 asked more detailed questions about the skills schools are building. The measures capture public judgment, not test performance. They are best read alongside—not as a substitute for—independent academic assessments.
The academic evidence is worrying, but it is not one-directional
PISA and NAEP tell a mixed recovery storyThe temptation in an anxious moment is to turn every assessment into a verdict. That is especially risky this week because two different systems—PISA and NAEP—measure different populations, use different frameworks and answer different questions.
PISA tests 15-year-olds across participating countries and economies and is designed around applying knowledge to real-world problems. In the United States, 4,625 students at 144 schools completed the 2025 assessment, representing roughly 3.58 million 15-year-olds in school. The OECD says U.S. reading performance was lower than in 2022, while mathematics and science were about the same as 2022. Reading and mathematics were below 2018 and among the lowest results the United States has recorded in PISA.
Yet the same country note reports that U.S. students remained above the OECD average in reading and science and close to the OECD average in mathematics. Seventy-four percent of U.S. students reached at least Level 2 in reading, compared with a 69% OECD average. Thirteen percent were top performers in reading, more than double the OECD average of 6%. In science, 78% reached at least Level 2 and 13% were top performers. In mathematics, 65% reached Level 2, matching the OECD average, while 8% were top performers, also matching the OECD average.
Those figures describe a system with substantial strength at the top and a serious problem lower in the distribution. Since 2015, the share of U.S. students scoring below PISA’s baseline Level 2 increased by seven percentage points in reading and six points in mathematics. The gap between the strongest and weakest readers is also exceptionally wide by international standards. That is a very different problem from “American students cannot read” or “American schools are globally uncompetitive.” It is a distribution problem as much as an average-score problem.
of U.S. 15-year-olds reached at least Level 2 in reading in 2025, above the 69% OECD average.
were top performers at Level 5 or higher, compared with 6% across the OECD.
increase in the U.S. share of low-performing readers below Level 2.
NAEP adds another layer. The 2024 main NAEP reading assessment found fourth-grade scores two points lower than 2022 and five points lower than 2019. Thirty-one percent of fourth-graders performed at or above NAEP Proficient, though that label should not be confused with “on grade level”; NAEP itself warns that its Proficient standard is a distinct, demanding performance benchmark.
For 12th graders, the 2024 NAEP reading score was three points lower than in 2019 and 10 points below 1992. That long slide is one reason concerns about adolescent literacy cannot be dismissed as a temporary pandemic aftershock.
Then came the June 2026 long-term trend results. Nine-year-olds improved four points in reading and four points in math from the most recent prior assessments. In reading, their 2025 average was not significantly different from 2020, meaning the younger cohort had essentially recovered to the pre-pandemic benchmark on that measure. Thirteen-year-olds did not show the same rebound: their reading and math averages were not significantly different from 2023 and remained below 2020.
The reasonable conclusion is not that recovery failed. It is that recovery is age-specific and uneven. Younger students have shown evidence of catching up. Older students who passed through critical years of schooling during the pandemic remain harder to move. Any national strategy that treats “learning loss” as a single uniform condition will miss that distinction.
A major PISA caveat for the United States
The OECD explicitly warns that the 2025 U.S. PISA results may be affected by participation and sampling problems. U.S. student response rates fell below the minimum target, exclusions exceeded the permitted tolerance, and school participation also fell short. The OECD says it cannot rule out bias or determine its likely direction. More than 40% of responding U.S. students also did not receive the background questionnaire, so the U.S. country note does not publish questionnaire-based findings. Performance results are informative, but they should not be treated as precision measurements free of sampling uncertainty.
What Americans say schools are not teaching well
Critical thinking and the transition to workGallup’s detailed skills questions help explain why a general “satisfaction” rating can fall even when many parents like their own schools. Adults are not only judging whether schools keep order or deliver diplomas. They are asking whether students are being prepared for an economy being reshaped by artificial intelligence, automation and rapidly changing entry-level work.
On the online survey, schools received their best marks for helping students adapt to new technologies, but even there only 44% of adults rated schools excellent or good. Thirty-two percent gave a positive rating for college preparation, 30% for effectively using AI tools and 28% for working effectively with others. Only 21% said schools were doing an excellent or good job preparing students for the world of work. Critical thinking ranked last at 20%, while 47% rated schools poor on that measure.
Those are perceptions, not direct measures of classroom skill. Still, they overlap with the themes in the new PISA framework, which emphasizes applying knowledge, evaluating information and solving unfamiliar problems. At the OECD-wide level, the 2025 results showed declines in the kinds of reading tasks that require students to connect information across sources and evaluate what they are reading. That is precisely the kind of capability adults increasingly associate with navigating AI-generated content, social platforms and a fragmented information environment.
There is also an important asymmetry. Schools receive their highest public rating for adapting to technology, but technology itself does not guarantee better learning. A classroom can be digitally modern and academically weak. It can also be deliberately low-tech in some moments and highly effective. The question is not how many screens or AI tools are present. It is whether students can use evidence, explain reasoning, write clearly, recognize unreliable claims and transfer what they know to a new problem.
Parents again offer somewhat better ratings than adults overall. Among current K-12 parents, 51% positively rate schools for preparing students to adapt to new technologies, 38% for working effectively with others, 37% for college readiness and 34% for using AI tools. But fewer than three in 10 parents give positive ratings for critical thinking or work preparation. The local-national gap narrows when the question becomes specific.
A school system can win credit for putting technology in students’ hands while still losing confidence over what students can do with information, judgment and real-world problems.
Why the public mood can fall faster than a school’s daily reality
Politics, expectations and uneven geographyNational satisfaction is partly a judgment about outcomes and partly a judgment about direction. Americans can see examples of strong schools and still believe the system is moving the wrong way. They can also approve of a child’s teacher while disapproving of district leadership, state policy, federal policy or the broader culture surrounding schools.
The Gallup responses show that the reasons for dissatisfaction differ by political identity. Republicans and independents were most likely to point to what they see as inadequate teaching of basic skills. Dissatisfied Democrats split more evenly among basic-skills concerns, insufficient funding or resources, and broader worries about education quality. Those are not the same policy diagnosis, even when they produce the same “dissatisfied” answer.
Federal policy has also become a visible part of the mood. The Trump administration has continued its effort to reduce the federal role and shift more authority toward states. On Sept. 9, the U.S. Department of Education released a “Learning Agenda Playbook” developed with eight state education agencies, presenting it as a framework for states to identify problems, build evidence and direct resources as more decision-making moves away from Washington. The department also used the week to emphasize evidence-based literacy initiatives.
Supporters of that direction argue that states and local communities can respond more directly to students and parents. Critics worry about capacity differences among states, uneven civil-rights enforcement, funding disparities and whether federal retreat will leave weaker systems with fewer tools. The new satisfaction numbers do not settle that argument. They instead raise the stakes: any governance model must show that it can improve student outcomes in ways families can see.
Geography matters, too. The United States does not operate one K-12 system. It operates thousands of school systems under different state standards, funding formulas, labor markets and demographic pressures. National averages can hide districts where reading recovery is strong and others where chronic absenteeism, staffing instability or poverty makes the school day far more fragile. The local-national satisfaction gap may therefore be partly rational: some parents really do live in places performing better than the national stories they read about.
The technology debate is moving from access to judgment
Devices, AI and what “future ready” should meanFor much of the past decade, the technology question in American schools was about access: which students had devices, broadband, digital curricula and modern learning platforms. The pandemic accelerated one-to-one laptop programs and pushed many districts to build a digital infrastructure almost overnight. By 2026, the harder question is what students should be doing with that infrastructure.
Gallup’s findings capture the shift. Adults are more confident that schools can help students adapt to new technology than that schools can help them think critically. That ordering matters in an AI economy. Knowing how to operate a tool is not the same as knowing when the tool is wrong, how to verify a claim, how to synthesize conflicting evidence or how to produce original reasoning without outsourcing the intellectual work.
PISA’s global findings point in the same direction, although the OECD warns that the United States lacks reliable questionnaire coverage for country-specific conclusions. Across participating systems, frequent AI use tended to be associated with stronger outcomes when students also had opportunities to develop AI literacy. The broader report also links stronger learning environments to teacher support and family engagement. Those relationships are not a prescription for a specific app or device policy. They are a reminder that technology works inside a human system.
This is why the next phase of school technology policy is likely to be less about “more” and more about boundaries. Districts are already debating when students should use screens, when paper is better, how to regulate phones, what AI assistance is permitted, what constitutes independent work and how teachers can redesign assignments so that a polished answer is not mistaken for genuine understanding.
The public’s low rating of critical-thinking preparation makes that debate urgent. If schools respond by layering AI products onto existing assignments without changing the intellectual task, they may reinforce the very concern the poll identifies. If they use AI as a reason to demand stronger source evaluation, oral explanation, drafting, revision, collaborative problem solving and transparent reasoning, technology could become a lever rather than a distraction.
Parents are not powerless observers of a national system
The local evidence families can actually inspectThe national debate tends to produce huge questions—Is public education failing? Is funding enough? Are standards too low? Is technology helping?—that no single family can answer. The more useful approach is to translate those questions into local indicators that can be observed over a school year.
Reading growth is one. Parents do not need to convert every assessment into a league table, but they can ask whether students who begin the year below benchmark receive targeted instruction, whether the school measures progress frequently enough to know if an intervention is working and whether older struggling readers have support rather than being treated as an elementary-only problem.
Attendance is another. Chronic absence is not simply a discipline statistic; missing instructional time makes every other improvement harder. Families can ask whether the school identifies absence patterns early, whether transportation or health barriers are involved and whether students returning after extended absences have a realistic academic recovery plan.
Teacher stability also matters. A school can have a strong curriculum on paper and still struggle if vacancies produce constant substitutions, large class loads or frequent course changes. Parents can look for continuity: Do students have qualified teachers in core courses? Are advanced courses consistently offered? Does the school retain staff long enough to build subject expertise and relationships?
Finally, families can ask what “future ready” means in concrete student work. Can students explain a mathematical choice? Can they summarize a difficult text without copying it? Can they distinguish an evidence-based claim from an unsupported one? Can they write and revise? Can they collaborate without one student carrying the group? Can they use an AI tool and then independently defend the answer it helped produce?
Five local questions worth asking this school year
- How is the school measuring reading growth for students who are below benchmark?
- What happens after repeated absences, and how quickly are families contacted?
- Where are the persistent teacher vacancies or high-turnover courses?
- How are phones, laptops and AI tools used differently by age and subject?
- What student work demonstrates critical thinking, revision and independent reasoning?
What would actually improve public confidence?
Results that are visible, comparable and hard to politicizePublic confidence will not recover because a federal agency, governor, superintendent, union or advocacy group declares that schools are improving. In a low-trust environment, messaging has limited power. Confidence is more likely to return when families see a pattern of results that survives political disagreement.
The first signal would be sustained literacy improvement across age groups, not only one rebound among younger children. The second would be narrower gaps between high and low performers. PISA shows the United States can produce many top readers and scientists; the challenge is raising the floor without lowering the ceiling. The third would be evidence that schools are preparing students for the transition to college, training and work—an area where public ratings are particularly weak.
Better data presentation would help, too. School accountability systems often overwhelm families with dashboards that are technically rich but practically inaccessible. A useful public report should make it easy to answer basic questions: Are students learning more than last year? Which groups are falling behind? Are absences improving? Are graduates completing demanding courses? How stable is the teaching staff? What support is offered when a student misses a benchmark?
The new federal Learning Agenda Playbook is built around a related premise: state agencies should define priority problems, identify evidence, test approaches and direct resources to what works. The value of that model will depend on implementation. A “learning agenda” becomes meaningful only when agencies publish questions that matter to students and families, use credible evidence, acknowledge failed approaches and change course when the data do not support a favored policy.
There is also a communications lesson in the Gallup local-national gap. Schools cannot assume that a family’s positive personal experience automatically translates into confidence in public education as an institution. Districts that want broader trust need to show how local successes connect to measurable outcomes—and they need to be equally candid about weaknesses. Trust built through selective good news is fragile. Trust built through useful information can survive a bad quarter or a disappointing test result.
Reading scores fell from 2022 and remained below 2019, while fourth-grade math improved.
Nine-year-olds improved; 13-year-olds showed no significant reading or math gain from their prior assessment.
U.S. adolescent reading weakens while national satisfaction with K-12 education falls to a new low.
The next school year will test whether this is a floor or a new normal
What to watch through 2026–27The most important near-term question is whether younger students’ NAEP recovery can be sustained as they move into middle school. A temporary bounce among nine-year-olds would be encouraging but insufficient. A durable recovery would appear in later grades, especially reading, where both domestic and international measures have shown persistent weakness.
Another test is whether states can translate the intense focus on literacy into consistent classroom practice. “Science of reading” policies, tutoring programs and curriculum changes have spread widely, but implementation varies. The relevant result is not how many laws or grants carry a literacy label. It is whether more students can comprehend increasingly complex text as they move from learning to read toward reading to learn.
Adolescent engagement will be equally important. The 2025 long-term NAEP findings show that older students have not recovered as strongly as younger children. Schools serving teenagers face a different problem set: course credits, attendance, phones and social media, part-time work, mental health, graduation pressure and the challenge of remediating foundational skills without making students feel as if they have been sent back to elementary school.
And then there is public trust itself. If satisfaction continues falling even where local outcomes improve, the country will have an information problem as well as an education problem. If local satisfaction begins to fall toward the national number, the warning becomes more serious: it would suggest that frustration is migrating from abstract perceptions into lived family experience.
For now, the most defensible reading is narrower. The United States has a genuine literacy and adolescent-learning challenge. It also has evidence of recovery among younger students, above-OECD-average performance in parts of PISA, a large population of high performers and parents who are still much more positive about the schools they know than the system they hear about.
That is not a comfortable conclusion, but it is a useful one. The country does not need a verdict about whether American education is “good” or “bad.” It needs a more disciplined way to identify which students are recovering, which are not, what local schools are doing differently and whether policy is producing results that survive the next election cycle.
Sources and methodology
- Gallup, “Satisfaction With K-12 Education Hits New Low,” Sept. 7, 2026 — national satisfaction trend, parent ratings and school-preparation measures.
- OECD, PISA 2025 Results: United States, Sept. 8, 2026 — U.S. performance, proficiency levels, sample description and data-quality cautions.
- National Center for Education Statistics, 2025 NAEP Long-Term Trend Results — reading and mathematics trends for 9- and 13-year-olds.
- National Center for Education Statistics, 2024 NAEP Reading at Grades 4 and 8 — recent reading score and achievement-level trends.
- National Center for Education Statistics, 2024 NAEP Reading at Grade 12 — high-school reading performance and long-term comparison.
- U.S. Department of Education, Learning Agenda Playbook announcement, Sept. 9, 2026 — current federal-state education policy context.
- Associated Press, Sept. 9, 2026 — independent reporting and context on the Gallup findings.
Comments
Post a Comment