UNESCO Institute for Statistics (UIS) indicator `FHLANGILP.LOWERSEC.HIGHSES` measures the percentage of lower-secondary students from a very affluent socioeconomic background whose first or home language is the language of instruction, for both sexes combined. The denominator is not all lower-secondary students. It is the subgroup classified as having a very affluent socioeconomic background, and the numerator is the students in that subgroup whose first or home language matches the language used for instruction.
The 2022 dataset contains 75 official national observations. The mean is 86.17% and the median is 93.50%. Twenty-seven observations are at least 95%, while six are below 50%. The maximum is 99.79%, followed by Japan at 99.24%, Colombia at 98.72%, Chile at 98.54%, and Brazil at 98.47%. At the lower end, the Philippines records 9.59%, Morocco 12.61%, and Brunei Darussalam 22.31%.
Table of Contents
What language match means in this indicator
The key phrase is ‘first or home language as language of instruction.’ The indicator asks whether the language a student first learned or uses at home is also the language used for teaching at school. A value of 90% means roughly nine in ten students in the specified socioeconomic subgroup are taught in a language that matches their first or home language.
First language and home language are not necessarily identical concepts in every household. Multilingual families may use several languages, and schools may teach in a national language, regional language, minority language, or another language. The UIS percentage summarizes this relationship for international comparison but cannot describe every linguistic arrangement inside a country.
The socioeconomic scope is deliberately narrow
The indicator applies only to students classified as having a very affluent socioeconomic background. It should not be rewritten as the national percentage for all lower-secondary students. Students from middle-income, disadvantaged, rural, migrant, or other socioeconomic backgrounds may have very different language experiences.
The phrase ‘very affluent’ is a category within the underlying socioeconomic measurement framework. The current dataset does not provide an absolute household-income threshold that can be applied uniformly across countries. It would therefore be misleading to translate the category into a specific income level.
The upper end is close to universal language match
The highest observations are clustered near 100%. The maximum is 99.79%, followed by Japan at 99.24%, Colombia at 98.72%, Chile at 98.54%, Brazil at 98.47%, Thailand at 98.46%, Mexico at 97.97%, El Salvador at 97.96%, Viet Nam at 97.90%, and Hungary at 97.89%.
There are 27 observations at 95% or higher. In these education systems, the vast majority of very affluent lower-secondary students in the observed sample have a first or home language that matches the language of instruction. The percentage does not show whether instruction is high quality or whether students achieve strong learning outcomes.
| Country or territory | 2022 language-match share |
|---|---|
| Republic of Korea | 99.79% |
| Japan | 99.24% |
| Colombia | 98.72% |
| Chile | 98.54% |
| Brazil | 98.47% |
| Thailand | 98.46% |
| Mexico | 97.97% |
| El Salvador | 97.96% |
| Viet Nam | 97.90% |
| Hungary | 97.89% |
The lower tail reaches below 10 percent
The Philippines records 9.59%, the lowest observation. Morocco is at 12.61%, Brunei Darussalam 22.31%, Malta 26.85%, the United Arab Emirates 36.51%, and Qatar 49.36%. Indonesia stands at 57.65%, Malaysia at 63.10%, Singapore at 71.52%, and Macao SAR at 72.28%.
Six observations are below 50% and another 5 are between 50% and under 75%. A low value means that a relatively large share of very affluent lower-secondary students are taught in a language different from their first or home language. The indicator does not by itself explain whether that pattern is related to multilingualism, migration, international schooling, historical language policy, or another factor.
| Country or territory | 2022 language-match share |
|---|---|
| Philippines | 9.59% |
| Morocco | 12.61% |
| Brunei Darussalam | 22.31% |
| Malta | 26.85% |
| United Arab Emirates | 36.51% |
| Qatar | 49.36% |
| Indonesia | 57.65% |
| Malaysia | 63.10% |
| Singapore | 71.52% |
| China, Macao Special Administrative Region | 72.28% |
Why the median is higher than the mean
The simple mean is 86.17%, while the median is 93.50%. The first quartile is 88.03% and the third quartile 95.85%. The middle half of observed countries and territories therefore lies between roughly 88.03% and 95.85%.
The mean is more than seven percentage points below the median because a small group of very low observations pulls the average downward. Values below 30% in the Philippines, Morocco, Brunei Darussalam, and Malta have a large effect on the arithmetic mean. The median better reflects the center of a distribution in which most observations are high.
The distribution is heavily concentrated above 90 percent
There are 6 observations below 50%, 5 between 50% and under 75%, 14 between 75% and under 90%, 23 between 90% and under 95%, and 27 at 95% or higher. Fifty of the 75 observations are therefore at least 90%.
That concentration should not obscure the magnitude of the lower-tail differences. The gap between the maximum and minimum is about ninety percentage points. Even within the same ‘very affluent’ socioeconomic category, the relationship between home language and school language differs sharply across national education and language environments.
Language match is not an enrolment or access measure
The denominator contains students already participating in lower-secondary education within the specified socioeconomic group. The indicator does not measure how many children of lower-secondary age are enrolled, how many are out of school, or how many complete the level.
Questions about access require separate indicators such as enrolment ratios, attendance, out-of-school rates, or completion. This series describes a characteristic of the instructional environment experienced by students who are in the assessed group.
Language match is not the same as learning achievement
Being taught in a familiar first or home language can be educationally important, but this percentage is not a reading score, mathematics result, or minimum-proficiency rate. A high language-match share should not be converted directly into a claim of better educational quality.
Learning outcomes require their own evidence. Completion, proficiency, assessment scores, and language-of-instruction alignment answer different questions. Combining them can be useful, but one should not be used as a substitute for another.
National averages can hide multilingual differences
A national percentage can conceal large differences between regions, linguistic communities, migration backgrounds, and school types. A country with a national value above 90% may still contain groups whose home language is rarely used for instruction, while a lower national value may contain regions with very high alignment.
The current dataset is national in geographic scope. It cannot identify which linguistic communities account for a high or low value. Regional or language-group data would be required for that level of analysis.
The affluent subgroup may not represent socioeconomic inequality
Very affluent families may have more school choice, live in different areas, or select schools with particular languages of instruction. The language-match percentage for this group cannot be used to conclude that wealth itself causes a higher or lower match.
To study socioeconomic inequality, the corresponding indicator for disadvantaged or middle socioeconomic groups would need to be compared with the affluent-group series. The current cross-country dataset only shows how the very affluent subgroup differs across countries.
Missing countries are not zero
The 2022 source contains 75 observed national rows. Countries and territories without a published observation are not systems with a 0% language match; they simply have no value in this indicator-year combination.
Zero-filling missing countries would distort the average, lower tail, and rankings. The correct analysis uses the 75 observed rows and treats unreported catalogue units as missing data.
What one year cannot tell us
The 2022 cross-section cannot show whether language alignment is increasing or decreasing. It cannot establish whether changes in migration, language policy, school choice, or the composition of the affluent subgroup are driving long-term trends.
Trend analysis should follow the same `FHLANGILP.LOWERSEC.HIGHSES` indicator across comparable assessment years. It is also important to check whether the socioeconomic classification and assessment coverage remain comparable across time.
Official source and interpretation
The official source is the UNESCO Institute for Statistics. This comparison uses 75 NATIONAL observations for 2022 from indicator `FHLANGILP.LOWERSEC.HIGHSES`, reported in percent for both sexes combined. It measures the percentage of lower-secondary students from a very affluent socioeconomic background whose first or home language is the language of instruction. Official values are available through the UNESCO UIS Data API.
The essential interpretive rule is to keep the population scope precise. This is a language-of-instruction indicator for the very affluent subgroup of lower-secondary students, not for all students, and it is not itself a measure of access or learning achievement. Within that scope, the data reveal very large cross-country differences.
Frequently Asked Questions
Does this indicator cover all lower-secondary students?
No. It covers only students classified as having a very affluent socioeconomic background.
Does a high language-match share mean high learning achievement?
No. The indicator measures language alignment, not test scores or minimum proficiency.
Are first language and home language always the same?
Not necessarily. Multilingual households can use multiple languages; the indicator checks whether a first or home language matches the language of instruction.
Can missing countries be treated as zero?
No. Countries without an official observation remain missing and must not be zero-filled.
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.
