What is the Flynn effect in simple terms?

The Flynn effect is the rise in average IQ test performance from one generation to the next that was measured across the 20th century — typically around 3 IQ points per decade. It is named after the New Zealand researcher James Flynn, who documented it in the 1980s. Because IQ tests are re-standardised so the average is always 100, the rise shows up when people take a test whose norms are older than they are.

What causes the Flynn effect?

No single cause is agreed, but the evidence points to the environment rather than genes, because the changes are far too fast for genetic change. The leading explanations are more and better schooling, better nutrition and health, smaller families and a daily life that demands more abstract, hypothetical thinking. A 2018 Norwegian study found that both the rise and the later decline appear between brothers in the same family, which rules out explanations based on which families have more children.

Are people getting dumber?

The data do not show that. Where scores have fallen, the drop is small — in Norway roughly 2 IQ points per decade for men born after 1975 — and much smaller than the rise that came before it. The Norwegian evidence places the cause in environmental changes within families, not in declining ability passed on from parents. Test scores also measure performance on particular tasks, which can shift with schooling and habits without any change in underlying capacity.

Which generation has the highest IQ?

IQ scores cannot answer that directly, because each generation is scored against its own norms and the average is set to 100 every time. On raw performance, people born in the late 20th century outscore their grandparents on the same tests by a wide margin. The US gain from 1932 to 1978 alone was 13.8 points. In countries where the trend has reversed, cohorts born after the turning point score slightly lower than those born just before it.

Why does the Flynn effect matter for IQ scores?

Because a test normed years ago compares you with a population that performed worse than today's, so your score comes out too high. A 2014 meta-analysis of 285 studies estimated this inflation at 2.31 points per decade, and about 2.9 points per decade for modern Stanford-Binet and Wechsler tests. That is why publishers re-norm their tests, and why a score is only meaningful against a recent comparison group.

Flynn Effect: What It Is, Causes and the Reversal

Q: Is the Flynn effect still happening?

It depends on the country. A 2015 meta-analysis of almost 4 million people in 31 countries found that the gains have become smaller in recent decades. In Norway, conscription records show scores rising for men born up to 1975 and falling for those born after. A US study of more than 390,000 adults tested online between 2006 and 2018 found declines on most reasoning item types but a rise on three-dimensional rotation.

The Flynn effect is the rise in average IQ test performance from one generation to the next — around 3 points per decade through most of the 20th century. It is named after James Flynn, who showed in 1984 that Americans had gained 13.8 points in 46 years. The rise has since slowed, and in several countries reversed — but the best evidence says both the rise and the fall come from the environment, not from genes.

What the Flynn effect is

IQ tests are standardised so that the average score of a reference group is 100. Every so often a publisher tests a new representative sample and resets the norms. In 1984 the New Zealand researcher James Flynn compared those successive samples and found that every Stanford-Binet and Wechsler standardisation sample from 1932 to 1978 set a higher standard than the one before. Put on a common scale, Americans had gained 13.8 IQ points in 46 years.

Three years later he collected data from 14 nations and found gains of 5 to 25 points in a single generation. The Dutch data were the most striking: of a 20-point gain, Flynn concluded that unknown environmental factors accounted for 15. The published average stayed at 100 throughout — only because the tests kept being re-normed. For how that scale is built, see what average IQ means and the IQ scale.

How big the gains were

The largest synthesis is a 2015 meta-analysis covering 1909 to 2013: 271 independent samples, almost 4 million people in 31 countries. It found the gains were not the same for every kind of test:

Type of test	Gain per year	Gain per decade
Fluid reasoning (new problems, patterns)	0.41 points	about 4 points
Spatial	0.30 points	about 3 points
Full-scale IQ	0.28 points	about 3 points
Crystallised (vocabulary, knowledge)	0.21 points	about 2 points

The same analysis found the gains were stronger for adults than for children, and smaller in recent decades. The pattern matters: the biggest rises came on the abstract reasoning tasks that were designed to be least dependent on schooling, not on vocabulary and general knowledge. The difference between those two kinds of ability is explained on our page about fluid vs crystallised intelligence.

Is the Flynn effect still happening?

Not everywhere. The clearest evidence comes from Norway, where most young men sat the same military ability test for decades. Scores rose for men born from 1962 to 1975 — about 0.26 IQ points a year within families, close to the global average — then turned. For cohorts born after 1975, scores fell, by about 0.23 points a year across the population between the 1975 and 1991 cohorts. Declines of this kind are often called the reverse (or negative) Flynn effect.

The United States has less data on adults. A 2023 study of 394,378 US adults who took the same open-source reasoning items online between 2006 and 2018 found a pattern consistent with a reversal: overall scores and matrix and letter-and-number-series scores fell, verbal reasoning barely moved, and three-dimensional rotation scores rose. An online sample is not a representative one, so the study describes a direction, not a national average.

One caveat pulls the other way. A 2014 meta-analysis of 285 studies found no sign of the effect weakening in comparisons between modern Stanford-Binet and Wechsler tests. The honest summary: the long rise is well established; its recent slowdown and reversal are real in some countries and some kinds of test, and the size of the change varies.

What causes it — and why genes are ruled out

Gains of 3 points a decade are far too fast to be genetic. The explanations researchers take seriously are environmental:

More and better schooling, lasting longer for more people.
Better nutrition and health, including fewer childhood infections.
Smaller families, with more attention per child.
A more abstract daily life. Flynn's own view was that modern life trains people to classify, to reason about hypotheticals and to handle symbols — the habits IQ tests reward.

How can something as heritable as IQ change so fast? A 2001 model by William Dickens and Flynn answers that environments and abilities amplify each other: a small advantage leads people into richer environments, which enlarge the advantage, and the same loop can run across a whole society. The finding that genes explain much of the variation between people does not stop the environment from moving the average — the argument is set out on our page on whether intelligence is genetic.

The Norwegian study tested the explanations for the decline directly. It compared brothers, who share their parents, and found that the rise, the turning point and the fall all appear within families. That rules out the popular claims that the decline comes from less intelligent parents having more children or from changes in who lives in the country. Whatever changed, it changed the environment the brothers grew up in.

Are people getting smarter, or just better at tests?

Flynn himself argued that the gains did not mean our grandparents were less intelligent in any everyday sense. In his 1987 paper he concluded that IQ tests capture something correlated with intelligence rather than intelligence itself — and pointed out that US scholastic aptitude scores fell during years when IQ scores rose. The gains were largest on the most abstract tasks, which is what you would expect if people became more practised at a particular way of thinking.

That also explains why the reverse Flynn effect does not mean people are becoming less capable. A fall in matrix-puzzle scores tells you that performance on matrix puzzles has fallen; it does not by itself tell you why, or what else has changed. Our page on how accurate IQ tests are covers what scores can and cannot show.

What it means for your IQ score

The practical consequence is about norms. If you take a test whose norms were set years ago, you are compared with a population that did worse than today's, so your score comes out higher than it should. The 2014 meta-analysis estimated the inflation at 2.31 points per decade across all tests and about 2.9 points per decade for modern Stanford-Binet and Wechsler tests. A test normed 20 years before you sit it can therefore add around 5 points. This is why clinical tests are re-normed regularly, and why a score from a test with old or unknown norms should be read with caution — the question is always “compared with whom, and when?” For how tests are built and normed, see how IQ tests work.

Trust and scope notes

This page explains published research in general terms. Population trends describe groups tested at different times; they say nothing about any individual's ability. Our IQ test measures reasoning and reports your result against its own norm group; it does not measure how your generation compares with another.

IQ Revealed is not affiliated with, endorsed by or connected to the researchers, universities, journals, publishers or test publishers named on this page, including the publishers of the Stanford-Binet and Wechsler tests. Their names are used only to identify the work being described.

The Flynn Effect: Why IQ Scores Rose — and Why They Are Now Falling