Average IQ by Country, Age, and Profession: What the Data Actually Says

Average IQ comparisons across countries, age groups, and professions all get thrown around on social media as though they’re equally solid. They’re not. Some rest on decent research; others rest on datasets that have been formally disavowed by the scientific community.

The short version up front:

• Average IQ by age is well-documented and reliable. The Flynn effect and cognitive aging curves have strong longitudinal evidence behind them.
• Average IQ by profession is reasonably well-documented with important caveats about selection effects.
• Average IQ by country is the murkiest. Almost all the numbers you’ve seen cited come from a single contested dataset that mainstream scientific bodies have advised against using.

Average IQ by country: the honest version

The IQ-by-country lists circulating online - the ones with specific numbers like “Hong Kong 108, Equatorial Guinea 59” - almost all trace back to a dataset compiled by psychologist Richard Lynn and collaborators, first published in IQ and the Wealth of Nations (2002) and updated in various forms, most recently by Lynn and David Becker in 2019.

The dataset has serious methodological problems:

• For 104 of the 185 countries in the original version, no actual IQ studies existed. Values were imputed by averaging neighbouring countries. El Salvador’s “IQ” of 84, for example, was calculated as the average of Guatemala (79) and Colombia (88).
• Much of the underlying data comes from small convenience samples - sometimes a few dozen subjects, sometimes only children, sometimes tested using instruments designed for different populations entirely.
• In July 2020, the European Human Behavior and Evolution Association issued a formal statement opposing use of the dataset, concluding that conclusions drawn from it are “unsound.”
• A 2022 analysis by Rebecca Sear, published in Frontiers in Psychology, found that updated versions inherit the same fundamental flaws. Publishers including Elsevier have since begun reviewing papers that used the data.

This doesn’t mean no cognitive differences exist between populations - it means the specific numbers attached to specific countries in these rankings don’t have solid empirical backing.

The cleaner alternative: PISA

A much more rigorous way to compare cognitive performance across countries is through international educational assessments like the OECD’s PISA (administered to 15-year-olds), TIMSS (mathematics and science), and PIRLS (reading). These tests use representative sampling, consistent instruments, and proctored administration. They’re not IQ scores, but they measure overlapping cognitive abilities and have solid statistical foundations.

Here are the highest-performing countries on the most recent PISA assessment (2022), by mean score in each subject:

A few things to notice. East Asian and northern European countries consistently cluster at the top - a pattern that shows up across PISA, TIMSS, and PIRLS, and is itself the source of a large research literature. The United Kingdom sits a little above the OECD average on all three. And crucially: these aren’t IQ scores. You can’t convert PISA to WAIS directly. But they’re the most rigorous cross-national cognitive proxy available.

Average IQ by age: the Flynn effect and cognitive aging

Across generations: the Flynn effect

The most robust generational finding in IQ research is the Flynn effect: average IQ scores rose by roughly 3 points per decade in most developed countries across the 20th century. A person scoring 100 on a 1950-normed test would score notably lower on a 2020-normed test - which is why test publishers like Pearson periodically re-norm their scales. The effect is named after researcher James Flynn, who documented it in the 1980s.

Causes are debated: improved nutrition, mass schooling, more abstract reasoning in everyday life (including, paradoxically, video games and screens), and better perinatal health are all candidates. No single explanation has consensus support.

Notably, the Flynn effect has stalled or partially reversed in several developed countries since around 2000 - including Norway, Denmark, Finland, and the UK. This has become an active research area, with possible causes ranging from changes in education to screen time to the declining quality of modern test norms.

Across your lifetime: cognitive aging

Within a single person, measured IQ is fairly stable after about age 16. But its components change in predictable ways across adult life. The main distinction:

• Fluid intelligence - reasoning with novel information, problem-solving, pattern recognition. Peaks in your late 20s and gradually declines from there.
• Crystallised intelligence - vocabulary, general knowledge, reasoning with familiar information. Keeps rising into your 60s or early 70s before slowly declining.

Aggregate IQ scores, which combine both, tend to peak in the 30s–40s and decline slowly afterwards, with decline accelerating in the 70s and 80s.

Individual variation is large. A well-educated, intellectually active 75-year-old routinely outperforms the average 30-year-old on many cognitive tasks, particularly anything involving accumulated knowledge or experience-based judgment.

Average IQ by profession

Occupational IQ data comes from a handful of well-cited sources: Hauser (2002), Gottfredson (1997, 2003), various military entry data (AFQT/ASVAB scores converted to IQ-equivalent scales), and large workplace samples from industrial psychology. The broad pattern is consistent across sources.

A few things to keep in mind reading this:

• These are medians, not requirements. A profession with median IQ 125 includes plenty of people at 115 and plenty at 135.
• They reflect selection effects, not job requirements. Engineering programs filter heavily on quantitative ability, which correlates with IQ; that’s why engineers have higher averages. It doesn’t mean you need 130 to be a good engineer.
• Most of these figures come from developed-world data from the 1990s–2000s. Contemporary and global patterns may differ.
• Within-profession variation is larger than between-profession variation for adjacent categories. A 90th-percentile accountant scores higher than a 10th-percentile physicist.

What these numbers can’t tell you

A few things worth being explicit about:

• None of these figures tell you anything about an individual. Knowing the average IQ of teachers is 112 tells you nothing about any specific teacher, who could score anywhere from 85 to 145.
• IQ isn’t fixed at birth. Education, childhood nutrition, and environmental factors all affect measured IQ, especially in childhood.
• High average IQ in a group doesn’t make low-IQ members of that group “smarter,” and vice versa. Averages are about distributions, not people.
• Cross-group IQ comparisons - national, ethnic, occupational - are frequently misused to support conclusions the data can’t actually support. Be especially skeptical of any source that jumps quickly from “group X has average IQ N” to “therefore...”

Frequently asked questions

The bottom line

What you can say with confidence: the Flynn effect exists, IQ varies across adult life in predictable ways, and occupations with strong educational filters have higher average cognitive scores.

What you can’t say with confidence: the specific “average IQ of country X is N” figure you’ve seen cited. If anyone tells you they know, ask for their source - and if the answer traces back to Lynn, Vanhanen, Becker, or The Intelligence of Nations, treat the number as folklore rather than data.