The Knowledge-Access Gradient: who can reach how deep into human knowledge, at what age, at what cost, and what that means for reforming education

The flagship synthesis of the education-atlas corpus. It consolidates the project's quantitative problem atlas, four structural deep-dives, four neutral foundations briefs, four solution-landscape briefs, and reproducible access analyses into one argument about the shape of knowledge access worldwide. The central finding: the world built a broad, shallow base of education and a thin, gated, unequal peak. Measured against UN SDG 4, three stacked crises persist: learning (48.3% of 10-year-olds cannot read a simple text; 86.5% in Sub-Saharan Africa), access (51M primary and 61M lower-secondary children out of school), and financing (3.6% of GDP, below the 4% floor). The deepest inequity sits on the depth axis. Laying knowledge access on a grid of age x depth (L0 basic literacy through L5 producing new knowledge), world access falls ~270x from undergraduate (37.4%) to the research frontier (0.14%), the rich-poor gap widens from under 2x at literacy to ~75x at the frontier, and about 99.86% of humanity only ever consumes knowledge while ~0.14% ever reaches the place where it is produced. Five expansion dimensions (cost, field, time, continuity, latency) show the binding constraint is production: a free path exists to READ the frontier and no path exists to PRODUCE there; coverage is wildly uneven across fields (biomedicine has ~133x more researchers than mathematics); reading is democratizing while production is flat; the pipeline leaks worst at the undergraduate-to-graduate step; and a learner faces ~13 gates to read the frontier and ~18 to produce there, 11 structural. The synthesis traces the causes (the industrial sorting model; a gatekeeping political economy in which academic publishers earn ~38% margins on donated labor; the omitted capacities of learning-to-learn, the body, and the ceiling on the ablest), and carries the corpus's foundations layer into the argument: what education is FOR is an irreducible, contested value question with no single cross-civilizational answer, and sorting, credentialing, and socialization are legitimate functions. Any reform is therefore a value choice and is stated as one. The evidence on what works (Teaching at the Right Level, tutoring ~0.37 SD, metacognition; learning styles null) and the white-space analysis converge on the L3-to-L4-to-L5 region as the region where expanding access to advanced knowledge pays off most. Bucket Foundation's open-knowledge thesis is named as one defensible choice among the contested aims and is bounded: it does not fix K-12 funding, teacher pay, or the access and learning emergency in poor states. Every headline number is traceable to World Bank EdStats, UNESCO UIS, OECD PISA, ITU, OpenAlex/research-atlas, or the peer-reviewed learning-science literature; constructed scales and estimated cells are flagged throughout. Part of Bucket Foundation's founding research for its education-reform mission. CC-BY-4.0.