PsychologyCounterintuitive resultDebated
Flynn Effect
Also known as IQ Rise, Intelligence Inflation
IQ scores rise 3 points per decade worldwide. If the tests measure intelligence, are we smarter? Or is something wrong with the tests?
- When
- 1984
- Difficulty
- Intermediate
- Reading time
- 7 min
Counterintuitive result. Nothing is wrong with the reasoning. The result is simply correct and human intuition is simply unreliable here. Not a contradiction at all.
Discover
To standardise an IQ test, you give it to a large representative sample and set the average score to 100. Then you publish it, and people take it.
A decade later, you give the test again to a new representative sample. If intelligence were stable in the population, they should still average 100.
They don't. They average about 103.
A decade after that, another group averages about 106. And so on.
James Flynn noticed this in the early 1980s while studying historical IQ norms. The numbers didn't add up unless scores had been rising. He collected the data from 14 nations and confirmed it: IQ scores had been rising in every country studied, by about 3 points per decade.
The gains are 30 points over a century. That is two standard deviations. If modern IQ test norms were applied to people from 100 years ago, approximately half of them would score below 70 — the traditional threshold for intellectual disability.
Does that mean our great-grandparents were intellectually disabled? Almost certainly not. So what does it mean?
Understand
The Flynn Effect is the steady, sustained rise in IQ test scores across generations, documented now in dozens of countries. The rise is approximately 3 IQ points per decade, though the rate varies by country and by period.
The gains are real in a narrow sense: people today perform better on the specific tasks tested by IQ tests than people a century ago. But what those gains mean is deeply contested.
What the gains cannot be. The rise is far too fast to be genetic. Genetic change in human populations operates over thousands of generations, not decades. 30 IQ points in 100 years — roughly two full standard deviations — cannot be explained by natural selection or any plausible genetic mechanism.
What the gains might be. Several environmental changes have been proposed:
Nutrition. Improved prenatal nutrition, reduced childhood malnutrition, and elimination of lead exposure (leaded petrol, lead paint) all affect brain development. The timing of the gains correlates with some of these improvements in some countries.
Education. Formal schooling, particularly with emphasis on abstract and analytic reasoning, may train the specific cognitive skills that IQ tests measure. Populations with more years of schooling tend to perform better.
Abstraction.* Flynn's own preferred explanation: modern life requires thinking in abstract categories — hypothetical scenarios, scientific logic, "what would happen if?" — in ways that earlier life did not. IQ tests measure abstract reasoning; the modern world generates more practice at it.
Test familiarity. Simply being more accustomed to tests, multiple-choice formats, and the style of questions asked may account for part of the gain.
What does this say about intelligence?
The Flynn Effect creates a philosophical puzzle. Either:
- People are genuinely more intelligent today than a century ago, in some meaningful cognitive sense.
- IQ tests measure something (performance on specific abstract tasks) that has improved, while the underlying intelligence they were designed to measure has not changed as much.
Most researchers favour interpretation (2). The gains are concentrated in the subtests furthest from everyday knowledge — abstract matrix reasoning, pattern completion, non-verbal spatial reasoning — and smallest in vocabulary and factual knowledge, where genuine cultural transmission would be expected to matter most. This suggests the tests are measuring something partly distinct from general intelligence.
The reverse Flynn Effect
Since approximately 1990–2000, researchers have documented that IQ gains have plateaued or reversed in several Nordic countries (Norway, Denmark, Finland), the UK, and Australia. The reversal is real, though smaller in magnitude than the original gains. Its causes are as disputed as the original effect's causes.
The shape of it
Same shape, different costume:
Examples
| Country | Period | Estimated IQ gain |
|---|---|---|
| United States | 1932–1978 | 14 points (Flynn's original study) |
| Netherlands | 1952–1982 | ~20 points |
| Israel | 1970–1990 | ~10 points |
| Sudan | 2010–2020 | 14 points (in a sub-study) |
| Norway | 1954–2002 | ~10 points — then reversal from ~1990 |
| UK | 1942–1992 | ~27 points (largest documented) |
The variation in magnitude is itself informative: the largest gains tend to occur in periods of rapid nutritional and educational improvement.
Explore
The abstraction hypothesis. Flynn's preferred explanation is cultural: modern people have learned to think in abstract,…
The abstraction hypothesis. Flynn's preferred explanation is cultural: modern people have learned to think in abstract, hypothetical categories that IQ tests require. He gives the example of a syllogism: "All bears are white. I saw a bear. What colour was it?" A pre-modern rural respondent might say "I've never seen a white bear" — treating it as an empirical question. A modern respondent says "white" — treating it as a logical exercise. The difference is not intelligence; it is familiarity with abstract reasoning as a mode.
Lead and IQ. One specific environmental factor with strong evidence is lead exposure. Leaded gasoline, phased out in most countries from the 1970s to 1990s, significantly raised blood lead levels in children in the mid-20th century. Blood lead levels above 5 µg/dL impair cognitive development. The phase-out of leaded petrol correlates with IQ gains in several countries, and some researchers estimate it accounts for a substantial fraction of the Flynn Effect gains, as well as the decline in violent crime since the 1990s.
What does reversing say? The reverse Flynn Effect in Nordic countries is most plausible for the same nutritional and environmental reasons that drove the initial gains: once lead is eliminated and nutrition is adequate, those environmental drivers are exhausted. What increases further may be largely test-taking practice; what decreases may reflect other environmental changes (screen time, changes in reading habits, social media).
Historical context
1932. The first large-scale IQ norming study in the US (the Army Alpha and subsequent intelligence tests). The baseline against…
1932. The first large-scale IQ norming study in the US (the Army Alpha and subsequent intelligence tests). The baseline against which future gains will be measured.
1940s–1970s. Regular renorming of IQ tests reveals that old norms produce inflated scores — a pattern that implies scores have been rising, though no one yet connects the data into a single narrative.
1984. James Flynn publishes "The Mean IQ of Americans: Massive Gains 1932 to 1978" in Psychological Bulletin. It is the first paper to document the gains systematically and frame them as a puzzle.
1987. Flynn extends the analysis to 14 nations in "Massive IQ Gains in 14 Nations: What IQ Tests Really Measure" in Psychological Bulletin, documenting that the phenomenon is cross-national.
1994. Richard Herrnstein and Charles Murray's controversial The Bell Curve acknowledges the Flynn Effect as a challenge to hereditarian views of intelligence — notable because the book is often read as supporting such views.
2000s–2010s. The reversal of the Flynn Effect is documented in several Nordic countries. Research on lead, nutrition, and schooling as specific mechanisms intensifies.
Why it matters
What intelligence tests measure. The Flynn Effect is the strongest argument that IQ tests measure culturally specific cognitive…
What intelligence tests measure. The Flynn Effect is the strongest argument that IQ tests measure culturally specific cognitive skills, not a culture-free general intelligence. If scores can rise 30 points over a century through environmental changes, the tests cannot be measuring something fixed and innate in any simple sense. This matters enormously for how we interpret IQ test results, what predictions we derive from them, and how we use them in education and policy.
Environmental determinism of cognitive development. The Flynn Effect is evidence that cognitive performance is substantially shaped by environment — nutrition, exposure to toxins, educational access, and cultural practices of abstract reasoning. This has policy implications: investing in these environmental factors, particularly in early childhood, produces measurable gains in cognitive performance.
Intelligence and heredity debates. The Flynn Effect is regularly invoked in debates about group differences in IQ test scores. If environmental factors can produce 30-point gains within populations over a century, they can also account for observed group differences, which are smaller in magnitude and associated with well-documented environmental differences.
Modern relevance
Lead and public health. The evidence linking lead exposure to cognitive impairment has renewed urgency as researchers have…
Lead and public health. The evidence linking lead exposure to cognitive impairment has renewed urgency as researchers have documented ongoing lead contamination in water pipes (the Flint, Michigan crisis), aging housing stock, and soil in former industrial areas. The Flynn Effect's lead story is not finished.
AI and cognitive demands. If the Flynn Effect reflects adaptation to the cognitive demands of modern life (abstraction, information processing, navigating complex systems), the increasing cognitive demands of digital life raise parallel questions. Are people developing new cognitive skills in response to smartphones and AI? Are they losing others (sustained attention, spatial navigation without GPS)?
Test design. Understanding the Flynn Effect has made test designers more cautious about what they claim their tests measure. The recognition that any test reflects the cognitive skills valued and trained by a particular culture — and that these change over time — is now standard in the psychometrics literature.
Sources & further reading
Primary source
(1984)The Mean IQ of Americans: Massive Gains 1932 to 1978Psychological Bulletin, 95(1), pp. 29–51.
The paper establishing the phenomenon in the US, documenting 14-point gains from 1932 to 1978 and ruling out genetic explanations.
How this connects
- Related to
- Dunning-Kruger EffectBoth concern the relationship between intelligence, self-assessment, and calibration. The Flynn Effect raises questions about what IQ tests measure; Dunning-Kruger raises questions about whether measured competence aligns with perceived competence.
- Goodhart's LawThe Flynn Effect can be read as Goodhart's Law applied to intelligence tests. As IQ tests became the target for educational systems, performance on them improved — raising the question of whether the tests still measure what they originally measured.