Are IQ Scores Falling?
The IQ Town Team
Psychometrics & test design
IQ scores are not falling in the simple sense headlines want. For much of the twentieth century, raw performance rose — the Flynn effect. In some countries, on some tests, that rise stopped or reversed. The labeled average is still 100, because testers reset the scale.
~3 pts
A typical decade of the old rise
1990s
When some countries reversed
100
The average after renorming
Not global
The reverse is uneven
The long rise
James Flynn and others showed that when people sat old tests, later generations outscored earlier ones. If you scored the old way, the 'average' person of the late twentieth century looked well above 100 by the norms of their grandparents. Better schooling, health, and familiarity with abstract test items are the leading explanations. Our page on the average IQ score already notes that the official average does not drift, because the test is renormed.
The partial reverse
From the 1990s, studies in Norway, Denmark, Finland, and a few other countries found raw scores flattening or declining, often by a couple of points a decade. The pattern is clearer on some subtests than others. It is not a uniform world decline, and it is not large enough to mean 'people became a different species of thinker.' Researchers argue about whether the easiest gains — basic schooling, nutrition, test-wiseness — had already been banked.
Years of school remain one of the few environmental factors with a clear link to scores. A reverse trend is a reason to look at those conditions, not a reason to announce a collapse.
The items change. This one is just a pattern.
Try one example before you start the full test. Instant feedback — no account required.
What number comes next in the series? 3, 6, 12, 24, 48, ?
Your score is not a generation's score
Take the free test and get a personal estimate, not a headline about a country.
How not to misread it
Do not read a national raw-score dip as your personal IQ, or as a ranking of peoples. This site's country pages use schooling as context for that reason: there is no official national IQ census. Do not read a chatbot era as the cause until someone measures it. How AI affects practice is a newer and smaller question than the Flynn curve.
Your own sitting is still the number that refers to you. Take the free IQ test if you want that, and ignore the decade-to-decade argument while you do.
The printed average cannot fall
As long as tests set the mean to 100, 'the average IQ fell to 94' is a confusion between raw performance and the scaled score.Frequently asked questions
Are people getting less intelligent?
In some countries, raw scores on some tests have slipped since the 1990s after a long rise. That is not a worldwide collapse, and the scaled average remains 100.
What is the reverse Flynn effect?
It is a stall or decline in the raw-score gains that defined the twentieth-century Flynn effect. It shows up in particular datasets, not as a single global fact.
Does a falling national trend change my score?
Only when the test is renormed, and then only by how you compare with the new group. Your performance on a given day is still your performance.
See where you land today
About 10 minutes, unassisted.