Education Services · UNESCO Institute for Statistics (UIS)

UIS Data Browser — UNESCO Institute for Statistics education statistics

Datadory delivers UIS Data Browser education statistics data covering the UNESCO Institute for Statistics' cross-country education corpus - 4,986 education indicators carrying 8,300,226 data points for more than 200 countries and territories, keyed one observation per indicator x country x year at national grain, spanning entrance ages, academic-year calendars, enrolment, out-of-school rates, literacy, teachers and education expenditure, with per-indicator histories reaching back as far as 1972 and up to 2025 in the February 2026 release. Get a sample of this dataset.

API, files, or your warehouse. Daily, weekly, or hourly.

Where it covers
More than 200 countries and territories reporting at national grain, plus SDG regional aggregates published as their own geounits
How far back
Per-indicator histories reaching as far back as 1972 and up to 2025; the current release is the February 2026 drop
How fine
Annual observations, one per indicator x country x year; no sub-national detail anywhere in the corpus
domains
Education carries 4,986 indicators and 8,300,226 data points; demography and socio-economics, science-technology-innovation and culture add 78 more

What is the UIS Data Browser education statistics dataset?

It is the widest table anyone keeps of who sits in a classroom, who pays for the seat, and who never gets near one. The UNESCO Institute for Statistics is the UN's official custodian of comparable national education statistics and the series behind SDG 4 monitoring, and its Data Browser is the interactive face of that corpus. The February 2026 release holds 4,986 education indicators carrying 8,300,226 data points, with 78 further indicators in three side domains: demography and socio-economics, science-technology-innovation, and culture.

Every measure is keyed to the same address - indicator x country x year, annual, national grain - whether the statistic is an entrance age, the month a tertiary academic year begins, an enrolment count, an out-of-school rate, a literacy figure or education spending. Regional aggregates ship as separate geounits rather than being blended into national rows, so a comparison of forty systems stays a filter rather than a reconciliation project.

What do sample records look like?

Two shapes, one system. Catalogue entries describe each indicator; observation cells hold its values:

# UIS catalogue entry - indicator 10, one of 4,986 education indicators
indicatorCode : 10
name          : Official entrance age to early childhood educational development (years)
theme         : EDUCATION
records       : 1,885 national observations
timeline      : 1972 - 2025

# observation cells - indicator 10405, France
indicatorId : 10405   geoUnit : FRA   year : 1970   value : 1969
indicatorId : 10405   geoUnit : FRA   year : 1971   value : 1970

Read the observation pair carefully and you learn why the dictionary matters: indicator 10405 records when the tertiary academic year starts, and its stored value arrives as a code rather than the words September. The flat shape never changes; the meaning of value travels with the indicator definition attached to it. Get a sample of this dataset

Which fields does the dataset carry?

Ten field families cover the whole corpus, and they divide cleanly into three jobs. indicatorCode, name and theme belong to the catalogue layer - what the statistic is and which domain it lives in. indicatorId, geoUnit and year form the address layer, pinning every number to one country-year cell. value is the payload, and magnitude and qualifier are the honesty layer riding beside it: precision flags and reporting qualifiers that tell you whether a figure is a national count, an estimate, or something odd enough to need a footnote.

lastDataUpdate closes the loop, dating each indicator's most recent refresh within its release. Everything beyond these ten - full geographic-unit reference lists, the complete indicator catalogue with per-indicator glossary text, the flag vocabularies - folds under additional fields on request.

Where does coverage run, and at what grain?

Geographically this is the widest net cast over world education: more than 200 countries and territories report at national grain, with SDG regional aggregates published alongside them as separate geounits. Where the OECD collections stop at member economies and partner countries, UIS continues down the income ladder until every system with a ministry and a school register has a row.

Temporally each indicator carries its own history - sampled timelines run from 1972 through 2025, and older series coexist with measures only a few reporting cycles old. The current vintage is the February 2026 drop, and working from the latest vintage matters because earlier years get revised as countries clean their returns.

Granularity is strictly one annual observation per indicator x country x year. There is no sub-national detail here at all - no provinces, no districts, no schools. Pair it with a national source when you need to descend below the country line; use it when the job is putting forty economies on one axis.

How does it compare to other international education datasets?

Think of the international education shelf as depth versus breadth. The OECD's education flows run deepest for its 38 members - teacher salaries, class sizes, graduate earnings - but stop at the membership line. **Eurostat's educ_* tables** give the EU27 and EFTA a tightly harmonised frame with the same ISCED ladder. Both are richer than UIS inside their borders and silent outside them.

UIS is the opposite trade: fewer analytical cuts per country, but every country on earth that reports, on the indicators the SDG framework actually monitors. When a board deck needs an out-of-school rate for markets the OECD has never heard of, or an impact report needs the citation every multilateral uses, this is the table. Most serious education-market models end up joining both - breadth from UIS, depth from the regional collections.

Field dictionary

Every field below is documented against real records. The full dictionary ships with the sample.

Field dictionary - UIS Data Browser education statistics (verified against live records)
FieldTypeDefinitionExample
indicatorIdstringNumeric UIS indicator code identifying the statistic - the join key between any observation and its definition.10405
geoUnitstringCountry or area identifier for the observation, ISO-like in form and resolved against UIS's geographic-unit reference list.FRA
yearintegerReference year of the observation. How far back a series reaches depends entirely on the indicator.1970
valuenumberObserved value of the indicator for that country-year. Read the indicator definition before reading the number: an entrance-age indicator stores years, a school-start indicator stores a coded month, a share stores its decimal.1969
magnitudestringMagnitude and precision flag attached to the observation under UIS metadata conventions. Blank in ordinary numeric rows.null
qualifierstringReporting qualifier describing how the value was produced - a national figure versus an estimate. Present as null in the sampled records.null
indicatorCodestringIndicator code in the definitions catalogue, pairing each statistic with its label and metadata.10
namestringHuman-readable indicator name carried alongside the code so rows stay legible without a lookup table.Official entrance age to early childhood educational development (years)
themeenumDomain grouping of the indicator: EDUCATION, DEMOGRAPHY_AND_SOCIO_ECONOMICS, SCIENCE_TECHNOLOGY_INNOVATION or CULTURE. 4,986 indicators sit in EDUCATION; the remaining 78 split across the other three domains.EDUCATION
lastDataUpdatedateWhen the indicator's figures were last refreshed, tagged with the release that carried the change.February 2026 Data Release

Coverage chips

DimensionCoverage
GeographyMore than 200 countries and territories reporting at national grain, plus SDG regional aggregates published as their own geounits
TimePer-indicator histories reaching as far back as 1972 and up to 2025; the current release is the February 2026 drop
GranularityAnnual observations, one per indicator x country x year; no sub-national detail anywhere in the corpus
DomainsEducation carries 4,986 indicators and 8,300,226 data points; demography and socio-economics, science-technology-innovation and culture add 78 more

What teams do with it

  • Sizing education markets country by country Enrolment counts by level, entrance ages, spending shares and pupil volumes define how many learners a market holds and at which stage. Rank expansion candidates by served population instead of GDP alone.
  • SDG 4 monitoring and grant reporting These are the canonical series the SDG 4 framework tracks, so progress reports cite the custodian's own numbers rather than a second-hand aggregation - and survive donor due diligence without a sourcing scramble.
  • Cross-country benchmarking decks One schema across 200+ economies ends the spreadsheet-stitching tax: identical indicator definitions, identical geography codes, regional aggregates kept separate from national rows. A forty-country pull is a filter.
  • Literacy and out-of-school targeting Out-of-school rates and literacy figures flag where unserved and underserved learner populations concentrate - the inputs an intervention funder or an edtech go-to-market plan needs before choosing territories.
  • Publishing and curriculum demand models Pupil volumes by education level, per country and per year, are the denominator behind textbook adoption, learning-material sales and assessment demand - feed them straight into unit-forecast spreadsheets.

Questions buyers ask

What is the UIS Data Browser education statistics dataset?

A structured extract of the UNESCO Institute for Statistics' cross-country education corpus, the internationally comparable series behind SDG 4 monitoring. The February 2026 release holds 4,986 education indicators with roughly 8.3 million data points for more than 200 countries and territories, plus 78 indicators across demography, science-technology-innovation and culture.

Which education indicators are included?

Everything from structural facts to outcome measures: official entrance ages by level, the month each academic year starts, enrolment counts, out-of-school rates, literacy, teachers, and public spending on education. Each indicator ships with its own definition, so a coded value such as a school-start month is never mistaken for a quantity.

How many countries does the dataset cover?

More than 200 countries and territories report at national grain, with SDG regional aggregates published as separate geounits. It is the broadest geographic reach of any international education collection - wider than the OECD and Eurostat frames, which stop at their memberships.

What years does the data span?

Each indicator carries its own history: sampled timelines run from 1972 through 2025, with older structural series reaching deepest. The current vintage is the February 2026 drop, and earlier years are occasionally revised within a release - another reason to work from the latest one.

What do the magnitude and qualifier flags mean?

They are the honesty layer on every observation. MAGNITUDE describes the nature and precision of the reading under UIS metadata conventions, and QUALIFIER records how it was reported - a national figure versus an estimate. Both were blank in the ordinary rows we sampled, which is itself informative: flagged cells are the exceptions worth pausing on.

Is the data comparable across countries?

That is the entire purpose. All values follow common UIS indicator definitions and a shared geographic-unit list, so comparing forty systems requires no reconciliation work. Honest limits remain: series length varies widely by country, and the flags exist precisely so a thin or modelled series can be weighted differently from a deep reported one.

How large is the dataset?

About 8.3 million education data points across 4,986 indicators in the February 2026 release - large enough to anchor production pipelines, small enough to profile whole in a notebook before you commit to a schema.

Can I evaluate the records before committing to a feed?

That is what the sample is for. Name the indicators you care about - enrolment, out-of-school rates, literacy, expenditure - and we cut sample rows toward them so you can validate country codes, value conventions and flags against your own pipeline first.

Datasets that pair with this one

See the rows before you pay anything.

Name this dataset and we send real records from it — scoped to the fields you asked for.

See pricing