The 2025 UNESCO Institute for Statistics (UIS) modelled series for male adolescents of lower-secondary-school age shows a very wide cross-country spread. Across 206 countries and areas, the median published out-of-school rate is 6.65%, while the simple unweighted mean is 13.66%. The mean sits well above the median because a relatively small upper tail reaches 40%, 50% and even nearly 70%. For this indicator, the distribution is more informative than a single average.
This is a male-only indicator. Each value represents the modelled percentage of males in the official lower-secondary-school-age group who are out of school under the UIS indicator definition. It does not directly compare boys with girls, and it is not a headcount. A sex-gap analysis would require the corresponding female series, while an estimate of the number of out-of-school boys would require the population of males in the relevant age group.
Table of Contents
A same-year comparison across 206 countries and areas
The official indicator code is ROFST.MOD.2.M. In the UNESCO UIS official API, all 206 rows used here are labelled 2025 and the unit is percent. That matters because this comparison does not mix older “latest available” observations with newer ones. The published values range from 0.0% to 69.9%, with a first quartile of 1.4% and a third quartile of 20.3%.
A common year does not mean the numbers explain their own causes. Cross-country estimates can reflect differences in education systems, underlying national data and the modelling process, among other factors. The data in this series identify the size of the estimated rate, not the mechanism behind it. For that reason, the discussion below describes levels and distribution without attributing a country’s value to income, conflict, policy, access or any other factor that is not present in the source data.
Why the 6.65% median is far below the 13.66% mean
When the 206 observations are ordered from low to high, the midpoint is 6.65%. The simple arithmetic mean is 13.66%, about 7.01 percentage points higher. This gap is a classic sign of a right-skewed distribution: many countries and areas have low single-digit rates, while a smaller set has very high values that pull the mean upward.
The 13.66% figure is not a global population-weighted out-of-school rate. Every country or area contributes one row, so a small territory and a very populous country receive the same weight in this descriptive average. A true global share of boys out of school would require the number of males in the official age group for every geography and an explicitly population-weighted calculation.
Most observations are below 10%, but the upper tail is substantial
Ninety-three countries and areas are below 5%, and another 29 are from 5% to under 10%. That places 122 observations below 10%. The middle and upper ranges are still large: 31 observations fall from 10% to under 20%, 14 from 20% to under 30%, and 19 from 30% to under 40%. Thirteen are from 40% to under 50%, while seven reach at least 50%.
| 2025 modelled rate | Countries/areas |
|---|---|
| <5 | 93 |
| 5–<10 | 29 |
| 10–<20 | 31 |
| 20–<30 | 14 |
| 30–<40 | 19 |
| 40–<50 | 13 |
| ≥50 | 7 |
These bands show why a single “typical country” is hard to describe. Nearly half the observations are below 5%, yet one quarter of the distribution begins around 20.3%. The lowest quarter and highest quarter therefore occupy very different parts of the scale. Reporting the quartiles and the band counts preserves that structure better than a rank list alone.
Seven published estimates are at or above 50%
The highest 2025 values are Niger at 69.9%, Equatorial Guinea at 65.3%, Burkina Faso at 57.9%, Liberia at 55.3%, the Central African Republic at 54.2%, Mali at 52.2%, and Afghanistan at 50.6%. Madagascar and the Syrian Arab Republic are both at 49.8%, and Chad is at 49.6%, just below the 50% threshold.
Those values are striking, but they should not be turned into a general ranking of national education systems. The indicator covers one sex, one official age group and one outcome: being out of school. Primary-school-age children, upper-secondary-age youth, girls, completion rates and learning outcomes can show different patterns. A country can therefore occupy a very different position when the question changes.
What a published 0.0% or 0.1% does—and does not—mean
Croatia, Ireland, Japan, Saint Kitts and Nevis, Georgia and Portugal have a published 2025 value of 0.0%. France, Montenegro, the Republic of Moldova, Lithuania and Saint Vincent and the Grenadines are at 0.1%. These are modelled percentages displayed at limited precision. A 0.0% entry should not automatically be rewritten as “zero boys out of school,” because rounding and modelling can produce a displayed zero even when the underlying population is not literally zero.
For the same reason, a difference such as 0.1 versus 0.2 percentage points should not be treated as a major substantive gap without uncertainty information and additional context. The more defensible comparisons in this table are broad ones: low single-digit values, rates above 10% or 20%, and the much higher observations above 40% or 50%.
How to read the “modelled data” label
The indicator name explicitly identifies the figures as modelled data. That means the published percentages should be treated as standardized estimates for international comparison, not as a single direct classroom census or a literal count of boys. A value of 20% does not mean that one survey physically observed exactly 20 out of every 100 boys as out of school. It is the rate UIS publishes for the indicator after its statistical production process.
This distinction is especially important when comparing countries with very close values. Modelled estimates are useful because they create a common analytical series, but they do not remove measurement limitations or provide causal explanations. The official rate is the starting point for comparison, while deeper country analysis needs the metadata, source inputs and complementary indicators.
The male series cannot answer the gender-gap question by itself
Because the series is explicitly male, it does not show whether boys are more or less likely to be out of school than girls in the same country. A male rate of 8% could coexist with a female rate of 3%, 8% or 15%, and each case would imply a different gender pattern. To discuss which sex has the higher rate, the female indicator for the same age group and year must be analysed alongside this one.
The same logic applies to parity language. A low male out-of-school rate is not evidence that gender parity has been achieved, and a high male rate does not tell us which sex is more disadvantaged. This dataset is strong for comparing the male group across countries; it is incomplete for comparing the sexes within countries.
What the 2025 snapshot can reveal
The clearest global pattern is the coexistence of many low values with a long upper tail. Ninety-three observations are below 5%, while 52 exceed 20%. Twenty are at or above 40%. The median-to-mean gap and the quartile spread tell the same story from different angles: the distribution is highly uneven across countries and areas.
What this one-year file cannot show is trend. It cannot tell us which countries improved from 2020 to 2025, whether a sharp change occurred in 2024, or whether a high 2025 value is persistent. Those questions require a multi-year series using the same indicator definition. A snapshot is valuable for level comparisons, but it should not be converted into a statement about direction of change.
Avoiding common comparison mistakes
There are several useful checks before interpreting the table. Confirm that the indicator is male-only, keep the unit as percent, and verify that all observations refer to 2025. Treat the values as modelled estimates, not raw counts. Do not turn the unweighted mean into a global population share. Do not infer causes from a one-variable table, and do not infer a gender gap from a male-only series. Finally, avoid overinterpreting tiny differences near zero where rounding and model precision matter.
Used with those limits in mind, the series provides a clear country-level picture of the estimated share of lower-secondary-age boys who are out of school in 2025. It is especially useful for identifying the broad distribution, the size of the upper tail and the countries with published rates far above the median, while leaving causal, trend and gender-comparison questions for additional data.
Frequently Asked Questions
What is the 2025 median male lower-secondary-age out-of-school rate?
The median across 206 countries and areas is 6.65%. The unweighted mean is higher at 13.66%, reflecting the long upper tail of high published rates.
Does a published 0.0% mean there are literally no out-of-school boys?
Not necessarily. The indicator is modelled and displayed at limited precision, so a published 0.0% should be read as a very low estimated rate rather than an exact zero headcount.
Can this dataset show whether boys or girls have the higher out-of-school rate?
No. It contains the male series only. A gender comparison requires the corresponding female indicator for the same age group and year.
Is 13.66% the global share of lower-secondary-age boys who are out of school?
No. It is a simple unweighted average of 206 country and area observations. A global population share would require age-specific male population weights.
Related Articles
Green Map creates custom-edited map images using open geographic data sources such as geoBoundaries, Natural Earth, OpenStreetMap, and government open data. These maps are edited visual materials, not raw data files, and are provided for education, documents, presentations, and graphic reference.
