Where "IQ by country" data comes from
Most public discussion of national IQ averages traces back to a single source: the dataset compiled by Richard Lynn and Tatu Vanhanen, published in books like IQ and the Wealth of Nations (2002) and updated in subsequent works.
Lynn and Vanhanen aggregated IQ test scores from studies conducted in different countries, often decades apart, often using different tests, and often with small or unrepresentative samples. They then computed average national IQ scores and published the resulting table.
The headline numbers vary somewhat by edition, but the overall pattern in their data shows highest scores in East Asia (~105), middle ranges in Europe and North America (~95–100), and lower scores in parts of Africa and South Asia. This dataset has been cited extensively — often by people drawing far stronger conclusions than the data supports.
Serious methodological problems
The Lynn-Vanhanen dataset has been heavily criticized in peer-reviewed literature. The main problems:
1. Inconsistent test administration
Different countries' data comes from different tests, administered under different conditions, in different decades, to different population subsets. Some figures come from samples as small as a few dozen people, often students or specific demographic groups rather than representative samples.
2. Cultural test bias
Many IQ tests were originally designed for Western, educated, industrialized samples. Items involving familiarity with specific objects, language, or scenarios can systematically disadvantage test-takers from different cultural backgrounds. Even "culture-fair" tests aren't fully culture-neutral.
3. The Flynn Effect
IQ scores have been rising across the world by roughly 3 points per decade throughout the 20th century — the Flynn Effect. This means older test data from one country, compared to newer data from another, will show a misleading "gap" that's really just a generation gap.
4. Estimation gaps filled by neighbors
For many countries, no actual test data existed. Lynn and Vanhanen filled gaps by using estimates from neighboring countries — a methodology that can create artificial regional clusters in the data.
5. Confounded with development
National IQ averages correlate strongly with national GDP, education systems, healthcare access, nutrition, lead exposure, parasite load, and infant mortality. Disentangling cause from effect is essentially impossible from observational data.
What actually causes measured IQ differences?
To the extent that measured IQ differences between populations are real (and not artifacts of methodology), well-documented causes include:
- Educational access and quality — Even one extra year of schooling can raise measured IQ by ~3 points.
- Nutrition — Iodine deficiency alone reduces measured IQ by 10–15 points in affected populations.
- Lead exposure — Childhood lead exposure causes measurable, permanent cognitive deficits.
- Healthcare and disease burden — Parasitic infections, malaria, and other endemic diseases impair cognitive development.
- Test-taking experience — Familiarity with standardized testing affects performance.
- Test culture-fairness — Some tests systematically disadvantage non-Western test-takers.
None of these is fixed or inherent. All can change — and have changed — when conditions change. The Flynn Effect is the clearest evidence: IQ scores rise within populations over time as conditions improve.
Within-country variation dwarfs between-country differences
Here's a fact that often gets lost: variation within any country is much larger than variation between countries. The standard deviation of IQ within a country is 15 points. Differences between national means are typically less than 10 points and often within methodological noise.
This means: pick any two countries at random. The odds that a randomly selected person from the "lower-IQ" country has a higher IQ than a randomly selected person from the "higher-IQ" country are usually around 35–45% — close to a coin flip.
National averages tell you almost nothing about any individual. Your neighbor, your colleague, your friend's IQ is essentially independent of their nationality.
The Flynn Effect — and what it means
James Flynn discovered that average IQ scores have been rising steadily worldwide throughout the 20th century, by about 3 points per decade. This effect is so robust that IQ tests have to be periodically re-normed to keep the population mean at 100.
This is direct evidence that measured IQ is highly malleable. If genes were the dominant factor in IQ differences between groups, large rapid changes within a single generation would be impossible. The Flynn Effect shows that environment — nutrition, education, exposure to abstract reasoning, even media — plays a huge role.
Recent data suggests the Flynn Effect may be slowing or even reversing in some developed countries, possibly as environmental gains plateau or new factors (screen use, education changes) push in different directions. The science here is still active.
How to read national IQ data responsibly
If you encounter national IQ averages — in headlines, on Wikipedia, in popular books — keep these things in mind:
- The numbers have large error bars. Don't treat 95 vs 102 as a meaningful precise difference.
- Within-country variation > between-country variation. Individual people are vastly more variable than national averages.
- Causes are environmental. Documented contributors include nutrition, education, healthcare, and lead exposure — all changeable.
- The Flynn Effect proves malleability. Whole populations' IQ scores rise across generations as conditions improve.
- Be deeply skeptical of conclusions about race, ethnicity, or "innate ability." The data can't support those claims regardless of how often someone repeats them.
For your own IQ, take a credible individual test. The number you get reflects your performance on specific cognitive tasks at this moment in your life — not your nationality, ethnicity, or any group identity.
