THE LONGEST-RUNNING FAILED EXPERIMENT IN AMERICAN HISTORY ISN'T A DRUG TRIAL — IT'S EDUCATION REFORM
How venture capital, Silicon Valley, and technocratic mandates keep failing students — and keep getting funded anyway
There's a famous definition of insanity, often misattributed to Einstein, that describes doing the same thing over and over while expecting different results. Whoever actually said it clearly never attended a high-powered education reform summit, because if they had, they'd have recognized the entire agenda. For roughly three decades, a rotating cast of billionaires, venture capitalists, and policy technocrats have been placing the same bet on public education — just with shinier packaging — and losing. Every. Single. Time. The students keep falling behind. The test scores keep dropping. And the next wave of "disruptors" keep cashing their checks.
Welcome to the most expensive, most confident, and most consistently wrong experiment in modern public policy.
The Factory Floor Model: Corporate Reform's Greatest Hits (and Misses)
The first wave arrived in the early 2000s with the swagger of a McKinsey consultant who just discovered a new PowerPoint template. The premise was elegant in its simplicity — and catastrophically wrong in its assumptions: run schools like businesses, measure everything, and competition will sort out the winners.
High-stakes standardized testing became the rectal thermometer shoved into education's mouth every year, regardless of whether the patient had a fever or was just nervous about the thermometer. Value-added teacher evaluations reduced the profound, complex art of human development to a regression coefficient. Charter networks expanded with the confidence of a franchise operation, promising that market forces would do what generations of dedicated educators apparently couldn't.
The results? Curriculum narrowed to whatever appeared on the test. Teachers burned out at industrial rates. Neighborhood schools — already underfunded — hemorrhaged resources to feed the accountability machine. And the equity gaps that corporate reform promised to close? They yawned wider.
The billionaires declared partial victory, pivoted their foundations, and moved on. The students stayed behind.
Silicon Valley Enters the Chat (And Never Leaves)
By the 2010s, a new savior had arrived, this time wearing a hoodie and carrying a pitch deck about "personalized, adaptive learning." The EdTech wave promised that algorithms could do what human teachers couldn't — meet every child exactly where they were, at exactly the right moment, with exactly the right content.
One-to-one device programs rolled out across districts with the enthusiasm of a product launch. Tablets replaced textbooks. Apps replaced worksheets. "Engagement" became the holy metric — and Silicon Valley, it turned out, was very good at engagement. Just not the kind that builds reading stamina, mathematical reasoning, or the capacity to sit with a difficult idea for longer than forty-five seconds.
What actually happened was almost poetically ironic: the same companies selling "personalized learning" platforms to schools were, after 3 p.m., deploying hyper-optimized social media algorithms specifically engineered to shatter the attention spans those same platforms were supposed to be developing. It was a vertically integrated attention destruction business, with the school day serving as the warm-up act.
OECD's PISA data has since told the story in cold, unambiguous numbers — continued declines in reading, math, and science literacy across high-income nations, with heavy device usage correlating directly with the sharpest score drops. Over a quarter of students across OECD nations now report frequent digital distractions during instructional time. The screens won. The students lost.
The Science of Reading: Genuine Research Meets the Venture Capital Machine
Here's where the story gets genuinely tragic, because the third wave — unlike its predecessors — is actually built on real science.
Cognitive neuroscience has spent decades establishing, with remarkable clarity, that young children learn to read best through explicit, systematic phonics instruction. This is not controversial among researchers. The "reading wars" between whole language and phonics advocates were, scientifically speaking, over. Phonics won.
But then the policy machine got hold of it.
State legislatures, eager to demonstrate action on declining literacy scores, began mandating "Science of Reading" frameworks with the same top-down confidence that had characterized every previous reform wave. And right on cue, the EdTech industry materialized with proprietary computer-adaptive testing software, scripted curricula sold by the seat license, and heavy platform contracts — all conveniently bundled under the "SoR" brand.
Settled cognitive research is now being rapidly converted into another profitable software pipeline. The science is real. The implementation is, in many jurisdictions, becoming a canned, scripted, one-size-fits-all product that strips teachers of the professional judgment that makes any pedagogical approach actually work.
It's the education reform equivalent of taking a genuine medical breakthrough and immediately turning it into a multi-level marketing scheme.
The Pattern That Never Changes
Across three decades and three distinct waves, critics and education researchers have identified a structural pattern so consistent it practically qualifies as a law of physics:
| Reform Wave | The Promise | The Actual Outcome |
|---|---|---|
| Market-Based Corporate Reform | Competition and accountability will drive excellence | Curriculum narrowing, teacher burnout, widening inequality |
| Ubiquitous EdTech | Personalized digital learning will engage every student | Distraction, attention erosion, declining literacy and numeracy |
| SoR Policy Mandates | Evidence-based literacy will rebuild foundational skills | Scripted software packages that undermine teacher autonomy |
The throughline is unmistakable: each wave treats complex human and pedagogical processes as engineering problems — as if children were buggy software awaiting a patch, and teachers were inefficient legacy systems awaiting replacement. Each wave arrives with philanthropic funding, legislative momentum, and an absolute certainty that this time the scalable, standardized solution will work.
Each wave leaves the same wreckage.
The Global Reckoning: OECD, UNESCO, and the Ministers Who Finally Noticed
The data has become impossible to ignore. The OECD's PISA findings now document what researchers have been warning about for years — a long-term structural decline in teenage reading, math, and science literacy across the world's wealthiest education systems. This isn't a pandemic blip. It's a trend line that predates COVID-19 and has been accelerating.
Global education ministers, to their credit, are beginning to respond — though with the characteristic speed of international bureaucracy, which is to say, urgently and about fifteen years late.
UNESCO's Digital Learning Week produced a joint ministerial framework that reads, in places, like a direct rebuke of the last two decades of EdTech enthusiasm. Its core pillars include:
- Preserving Cognitive Agency — explicitly warning against AI tools that bypass critical thinking and original problem-solving
- Protecting Teacher Autonomy — ensuring educators retain final authority over which tools enter their classrooms
- Data Privacy Safeguards — requiring that student data remain non-commercialized and auditable
- Equity and Anti-Bias Provisions — auditing educational AI for cultural and linguistic bias, championing open-source alternatives
Meanwhile, education ministries across high-income nations are implementing device bans, pivoting back to structured literacy and explicit math instruction, and — in a development that would have seemed radical five years ago — recommending that schools put the physical books back.
The policy responses are, in other words, largely a return to what worked before the disruption cycle began. Which raises an uncomfortable question that no one in the reform industry seems eager to answer.
What Actually Works (And Why It's Never Been the Interesting Answer)
The research on sustainable educational improvement is, at this point, fairly consistent. It is also, notably, not very fundable as a venture investment, because it doesn't scale into a software platform.
Reduce class sizes. Smaller groups allow teachers to provide the responsive, relational instruction that no algorithm has successfully replicated. It requires hiring more teachers, which costs money and cannot be automated.
Pay teachers competitively. Teacher shortages are now a critical policy crisis across high-income nations. Addressing them requires treating teaching as the skilled, cognitively demanding profession it is — not as a delivery mechanism for a scripted curriculum.
Restore professional autonomy. The evidence consistently shows that teacher expertise, exercised with genuine professional judgment, outperforms scripted, standardized delivery. This means trusting teachers, which is philosophically incompatible with the technocratic reform model.
De-center screens. Print-rich environments, physical books, collaborative group work, and tactile learning experiences build the sustained attention and deep reading stamina that screen-based learning has demonstrably eroded.
Address out-of-school factors. Poverty, housing instability, and inadequate healthcare access account for a substantial portion of the score disparities that no app, no testing regime, and no scripted curriculum can fix. These are policy problems, not pedagogy problems.
None of these solutions require a platform. None of them generate recurring subscription revenue. None of them can be pitched at a TED Talk with a compelling slide deck showing exponential growth curves.
They are, however, what the evidence supports.
The Billion-Dollar Lesson Nobody Wants to Learn
The "Reading Wars" industry alone has been estimated at over a billion dollars — a remarkable figure for a debate that cognitive science resolved decades ago, and an even more remarkable figure given that the resolution has yet to meaningfully improve outcomes at scale.
The broader corporate education reform enterprise has consumed hundreds of billions in public funds, philanthropic capital, and venture investment over three decades. The OECD data suggests that students in high-income nations are, by measurable outcomes, worse off than when the disruption cycle began.
This is not a failure of effort. The reformers have been relentlessly energetic. It is not a failure of funding. The money has been extraordinary. It is a failure of model — a fundamental category error that mistakes education for a product delivery problem, and students for consumers awaiting a better user experience.
Public education is not a market. Children are not customers. Teachers are not inefficient intermediaries awaiting disintermediation. And learning — the deep, slow, relational, cognitively demanding process of building a human mind — has never responded well to being treated as a software deployment challenge.
The billionaires will, of course, be back with the next wave. The pitch deck is probably already in production. The brand will be different. The underlying assumption — that the right technology, the right metric, the right scalable intervention can finally crack the code — will be identical.
And somewhere, a classroom teacher will be doing what classroom teachers have always done: building relationships with students, making a thousand micro-decisions per hour based on professional expertise, and quietly producing the kind of learning that never shows up cleanly in a venture fund's impact report.
The insanity, it turns out, was never in the students. It was always in the funding model.
The OECD PISA international skills assessment and UNESCO's Digital Learning Week ministerial framework both reflect ongoing global policy responses to documented declines in student learning outcomes across high-income nations.
Sources & References
🔴 OECD PISA International Skills Decline
1. OECD — PISA 2025 Press Release: Students' Reading and Mathematics Performance Declined Sharply Across the OECD
"Average reading scores fell by 28 points across OECD countries between 2015 and 2025, equivalent to about one and a half years of learning." 🔗 https://www.oecd.org/en/about/news/press-releases/2026/09/pisa-2025-students-reading-and-mathematics-performance-declined-sharply-across-the-oecd.html
2. OECD — PISA 2025 Results (Volume I): Full Publication
Mathematics scores fell by 22 points between 2015 and 2025 across OECD nations. 🔗 https://www.oecd.org/en/publications/2026/09/pisa-2025-results-volume-i_5265bfb1.html
3. OECD — PISA Programme Overview 🔗 https://www.oecd.org/en/about/programmes/pisa.html
🏭 Corporate Education Reform & Billionaire Philanthropy
4. Dissent Magazine — "Got Dough? How Billionaires Rule Our Schools"
Hundreds of private philanthropies spend nearly $4 billion a year transforming K–12 education, led by Gates, Broad, and Walton foundations. 🔗 https://dissentmagazine.org/article/got-dough-how-billionaires-rule-our-schools/
5. Institute for New Economic Thinking — "Millionaire-Driven Education Reform Has Failed. Here's What Works"
Journalist Andrea Gabor's analysis of why top-down philanthropic reform has consistently underdelivered. 🔗 https://www.ineteconomics.org/perspectives/blog/millionaire-driven-education-reform-has-failed-heres-what-works
6. Public Integrity — "Grading the Billionaires on Education Reform"
An accountability audit of major philanthropic education investments and their measurable outcomes. 🔗 https://publicintegrity.org/accountability/grading-the-billionaires-on-education-reform/
7. ResearchGate / Academic Paper — "The End of Public Schools? The Corporate Reform Agenda to Privatize Education"
Peer-reviewed analysis of how corporate reform models have restructured public education policy. 🔗 https://www.researchgate.net/publication/318030875_The_end_of_public_schools_The_corporate_reform_agenda_to_privatize_education
📱 EdTech, Screen Time & Student Distraction
8. Science News — "Science Backs Calls for Limiting Screens in Schools"
EdTech use linked with declines in test scores in the U.S. and abroad, consistent with multiple peer-reviewed studies. 🔗 https://www.sciencenews.org/article/science-school-screen-limits-devices
9. Ethics and Public Policy Center (EPPC) — "Getting Screens Out of Schools: EdTech Screens Are Hurting, Not Helping Student Learning"
Comprehensive policy analysis documenting the correlation between device proliferation and declining math and reading scores. 🔗 https://eppc.org/publication/getting-screens-out-of-schools/
📖 Science of Reading & Corporate Capture
10. Nancy Bailey's Education Website — "The Science of Reading Corporate Connection: Replacing Teachers with Tech"
Detailed critique of how SoR policy mandates are being bundled with scripted, vendor-driven curricula that undermine teacher professional judgment. 🔗 https://nancyebailey.com/2022/05/23/the-science-of-reading-corporate-connection-replacing-teachers-with-tech/
11. London School of Economics — "In the Fine Print: Investigating EdTech Providers' Data Protection Practices"
Working paper examining data privacy risks embedded in EdTech vendor contracts deployed in K–12 schools. 🔗 https://www.lse.ac.uk/social-policy/Assets/Documents/PDF/working-paper-series/WPS-02-25-EdTech.pdf
🗂️ Quick Reference Summary Table
| **# ** | Source | Topic |
|---|---|---|
| OECD PISA 2025 Press Release | Skills decline data | |
| OECD PISA 2025 Volume I | Math/reading score drops | |
| OECD PISA Programme Page | Assessment overview | |
| Dissent Magazine | Billionaire philanthropy | |
| INET Economics | Reform failure analysis | |
| Public Integrity | Billionaire accountability audit | |
| ResearchGate Academic Paper | Corporate privatization critique | |
| Science News | Screen time & score declines | |
| EPPC Policy Report | EdTech harm in classrooms | |
| Nancy Bailey Education | SoR corporate capture | |
| LSE Working Paper | EdTech data privacy risks |
All links were verified as active and relevant as of September 2026. The OECD PISA 2025 sources ( and ) are the most current international data available, published this month.