The Smartest Country in the World: The Winner Nobody Expected

Ask the internet for the smartest country in the world and you get a clear answer.
Japan. 106.48.
Second place goes to Taiwan, at 106.47.
The gap between first and second is 0.01 points.
A gap of 0.01 on a scale where one person can score a few points higher or lower on a different day is not a gap at all. It is a rounding mark.
Japan sits at the top of this table, yes. But the honest reading is that the top two are tied.
This post looks at the winner alone. Why Japan is there. What the number is made of. And what it does not mean.
One thing first. The title says the winner is nobody’s guess. That is the hook, not the fact.
On this dataset, Japan is the expected winner. It tops the table in every edition of these rankings. The surprise is not who won. The surprise is how small the win is.
The gap that is not a gap
An IQ score has a middle of 100 and a spread of 15 points. That spread is the unit the whole scale is built on.
A single person’s score also carries error. Take the same test after a bad night of sleep, and the number can move by a few points.
Now set 0.01 beside that.
The gap between Japan and Taiwan is one hundredth of a point. It is far smaller than the wobble on any one person’s score. It is smaller than the wobble on any small group of people.
So what does “Japan wins” mean here?
It means the maths put Japan first by a hair. The press release for the 2019 book said it more honestly: Japan, Taiwan and Singapore share the top spot, all at about 106.
Call it a photo finish. Three runners cross the line together, and the camera gives the win to one of them.
That is the first thing to hold on to. The number is real, in the sense that it comes from real test data. The win is not real, in the sense of a clear distance between two countries.
Why Japan sits at the top
So why is Japan there at all, year after year, edition after edition?
The short answer is school, health and practice with tests.
Japan has near total literacy. Nearly every child finishes school. Maths is taught with care from an early age, and test taking is a normal part of school life from childhood onward.
Food and health help too. A child who eats well and sleeps well will test better than a child who does not.
Japan also has a lot of test data behind it: many studies, over many years, with large groups of people. A country with fifty studies behind its number rests on firmer ground than one with a single small study.
None of this proves that one nation is born smarter than another. It shows what a country looks like when most of its children get long schooling, good health and years of practice with the exact skills these tests measure.
That is a story about systems. Not about blood.
The top ten, in short
Here is the table this post is built on, from Lynn and Becker 2019. It is a short recap only. The full ranking, with each place explained, was covered in an earlier post.
1. Japan, 106.48
2. Taiwan, 106.47
3. Singapore, 105.89
4. Hong Kong, 105.37
5. China, 104.10
6. South Korea, 102.35
7. Belarus, 101.60
8. Finland, 101.20
9. Liechtenstein, 101.07
10. Netherlands and Germany, 100.74
The gap from first to tenth is 5.74 points. On a scale with a spread of 15, that is less than half a spread across ten whole countries.
The top five are all in East Asia. The scores step down gently, a point or two at a time. There is no cliff, and no runaway leader. A table this tight should make a reader cautious, not confident.
What the number is made of
The source is a book. Richard Lynn and David Becker, The Intelligence of Nations, published by the Ulster Institute in 2019.
It is the latest version of a project Lynn worked on for decades. The 2019 dataset pulls together 667 studies from around the world, and turns them into one number per country.
Six hundred and sixty seven studies is a lot of data.
But look at how the number is built, and the solid ground gets soft.
The studies are mixed. Different tests. Different age groups. Different years, some of them very old. Different ways of picking the people who took the test.
For countries with no study at all, the book fills the gap with a guess based on nearby countries. That is written in the method, not hidden. It is still a guess.
So 106.48 is not a measurement, like a height taken with a tape. It is an estimate, built by averaging a pile of mixed studies and filling the holes with neighbours.
Estimates can be useful. But two decimal places do not make a number precise. They only make it look precise.
What this number does not show
Here is the caveat section, and it is a big one, because this dataset has drawn serious criticism from other researchers.
A peer reviewed paper by Sear, Lawson, Kaplan and Shenk in 2022 looked at the 2019 dataset and found it unfit for the job it is used for.
The details are plain. About 37 in every 100 samples had fewer than 1,000 people in them, and about 74 in every 100 had fewer than 5,000. About 63 in every 100 tested only people under 18, and about 30 in every 100 countries rest on a single sample. Six countries rest on fewer than 100 people.
A sample of 34 people cannot stand for a whole country, and a sample of school children cannot stand for adults.
Then there is the Flynn effect. Scores rose by about 3 points a decade through the last century, as schooling and daily life changed. A study from 1970 and a study from 2010 are not measuring the same world.
Then there is the test itself. Most IQ tests were built in the West, in Western languages, around Western school skills. A test built in one culture will always favour the children who grew up inside it, and no amount of averaging removes that bias.
So what is a better yardstick? School tests like PISA are a fairer start. PISA gives the same test, to fifteen year olds, in the same year, in every country that takes part. PISA 2022 tested about 690,000 students in 81 countries and economies.
Singapore came top in all three subjects: maths 575, reading 543, science 561. Japan sat near the top too: 536 in maths, 516 in reading, 547 in science.
PISA is not an IQ test. It measures school learning, not raw ability. But it is the same test, given the same way, to the same age group, and that is a kind of fairness the Lynn and Becker numbers cannot offer.
And one more fact, the most important one in this post.
A national average says nothing about any one person in that country.
Japan’s average is 106.48. Scores in Japan spread out just like scores everywhere else. Millions of Japanese people score below 100. Millions of people in countries far down the table score above 130.
An average describes a crowd. It never describes the person standing in front of you.
Wrapping up: a photo finish, not a crown
Japan tops the table. That is what the data says, and there is no reason to hide it.
But the win is 0.01 points wide, on a scale where a bad night of sleep can move a score by more than that. The top two are tied in every way that matters. The top ten sit within six points of each other.
Read the number as a rough sketch of schooling and test history, not as a verdict on a nation. If you want to know how sharp a country’s young people are, look at a test they all sat on the same day, like PISA.
The smartest country in the world, on this dataset, is Japan by a rounding mark.
The smarter question is what the number is worth. And the answer to that one is: less than the headline suggests.
