The Flynn effect is the long-term rise in average IQ test scores seen across the 20th century in dozens of countries, typically around 3 points per decade. It means that a person scoring 100 on norms from 1950 would score well below 100 against norms from 2000. Researchers agree the rise was real and mostly environmental. They disagree about how much of it reflects a true increase in general intelligence, and in several countries, including Norway and Denmark, scores have flattened or fallen for people born since the 1970s and 1980s.
What is the Flynn effect
Because IQ tests are scaled so that the average of a norm sample is 100, any rise in raw performance is hidden until someone compares groups against the same yardstick. The New Zealand–based political scientist James Flynn did exactly that. In 1984 he showed that when American test-takers from different decades were scored against the same norms, performance had risen by about 3 IQ points per decade between 1932 and 1978 (Flynn, 1984).
In 1987 he gathered data from 14 nations and found gains everywhere he looked, from about 5 to 25 points in a single generation. The biggest gains were on nonverbal reasoning tests such as Raven's Progressive Matrices, which were thought to be the least dependent on schooling. Dutch military conscripts, for example, gained about 20 points on a Raven's-type test between 1952 and 1982 (Flynn, 1987). Rises had been noticed earlier, but Flynn showed how large and widespread they were, and the authors of The Bell Curve later gave the phenomenon his name.
How big was the rise
The most comprehensive estimate comes from a meta-analysis by Jakob Pietschnig and Martin Voracek covering 271 independent samples, nearly 4 million participants and 31 countries from 1909 to 2013. They found consistent gains, roughly 3 points per decade for full-scale IQ, larger for fluid reasoning than for crystallized knowledge, and a slowing of the gains in more recent decades (Pietschnig & Voracek, 2015).
Taken at face value, a century of gains at that rate adds up to around 30 points. That would put the average person of 1910 at about 70 by today's norms, a score now associated with intellectual disability. Nobody seriously believes most people in 1910 were intellectually disabled. That paradox is what makes the effect so important for understanding what IQ tests measure.
Are people getting smarter
People today are clearly better at the kinds of problems IQ tests pose, especially abstract classification and pattern reasoning. Whether that means they are smarter depends on what you mean by smart.
Flynn's own view was that modern life trained people to think in abstract, hypothetical categories: to answer "what do a dog and a rabbit have in common" with "both are mammals" rather than "you use a dog to hunt rabbits". Our great-grandparents were not less capable; their thinking was tuned to concrete, practical problems, while ours has been shaped by schooling, office work and a world full of symbols and classifications.
Statistical work supports the idea that the gains are not simply more of the same ability that separates individuals within a generation. When researchers tested whether IQ batteries measure the same constructs across cohorts, they found that the tests did not behave identically over time, meaning score gains between generations cannot be read in the same way as score differences between people in one generation (Wicherts et al., 2004). In other words, the Flynn effect is real, but it is not straightforward evidence of an equally large rise in the g factor.
What caused the Flynn effect
The rise happened far too quickly for genetic change to explain it, so researchers focus on environmental causes. No single explanation accounts for the whole effect, and the likely mix differs by country and era.
| Proposed cause | Why it is plausible | Limits |
|---|---|---|
| More and better schooling | Years of education rose sharply during the century, and education has a measurable effect on IQ | Gains were largest on tests meant to depend least on school knowledge |
| Better nutrition and health | Height and health improved alongside scores; childhood illness and malnutrition impair cognitive development | Gains continued in rich countries after nutrition was already good |
| A more cognitively demanding environment | Work, media and daily life increasingly reward abstract reasoning | Hard to measure directly |
| Smaller families | More parental attention per child | Probably a minor contributor |
| Greater test familiarity | Testing and test-like formats became common in schools | Unlikely to explain gains of this size alone |
William Dickens and Flynn proposed a mechanism that ties these together. Small environmental advantages lead people to seek out more stimulating experiences, which raise ability further, and as a whole population shifts, each person's surroundings become more demanding too. This "social multiplier" can let modest environmental changes produce large gains in scores, while genes still explain a large share of differences between individuals within a generation (Dickens & Flynn, 2001). That reconciles the effect with high heritability estimates, which describe variation within a population, not changes over time; see nature vs. nurture in intelligence.
The reverse Flynn effect
Since the 1990s, several countries with long-running conscription testing have reported that scores stopped rising and then began to fall. In Denmark, scores on the conscription test peaked in the late 1990s and declined in the following years (Teasdale & Owen, 2005). Similar reports have come from Norway and a few other countries, mostly in Northern Europe.
The strongest study comes from Norway. Bernt Bratsberg and Ole Rogeberg analyzed about 730,000 IQ scores from Norwegian men tested at conscription, born between 1962 and 1991. Scores rose for those born until the mid-1970s and then declined. Crucially, the decline also appeared within families: younger brothers tended to score lower than their older brothers in the later cohorts. That rules out explanations based on which parents were having more children or on changes in who was being tested, and points to environmental causes (Bratsberg & Rogeberg, 2018).
Which environmental changes are responsible is still unknown. Proposals include changes in education and reading habits, media use, and shifts in what tests measure relative to what young people now practice. None of these has been established.
Is there a reverse Flynn effect in the United States
The evidence is mixed. A study of a large online sample of US adults tested between 2006 and 2018 found declines in several areas, including verbal reasoning and matrix reasoning, alongside gains in three-dimensional spatial reasoning (Dworak et al., 2023). The sample was self-selected rather than representative, so it is a signal to watch rather than a settled finding. It is fair to say that the steady American gains of the 20th century can no longer be assumed.
What the Flynn effect means for IQ test norms
The most practical consequence is that test norms age. If average performance rises about 3 points per decade, a test normed 15 years ago will give today's test-takers scores several points too high, because they are compared with an older, lower-scoring reference group. This is why publishers re-norm their tests every 10 to 20 years: the WAIS-5, released in 2024, replaced WAIS-IV norms collected in 2007–2008. It is also why you should always check which test and which norms produced a score. Our article on how IQ scores are calculated explains how norms work.
The stakes can be serious. A diagnosis of intellectual disability uses scores around 70 as one criterion, so a few points of norm inflation can change whether someone qualifies for services. In US death-penalty cases, where the Supreme Court in Atkins v. Virginia (2002) barred executing people with intellectual disability, experts and courts have argued over whether to adjust scores from outdated norms. Practice varies, and this is an area for qualified clinicians and lawyers, not for general rules.
The effect also complicates comparisons across time and place. Average scores from different decades are not directly comparable unless they were scored on the same norms, and comparisons between countries are affected by differences in schooling, health and test familiarity, along with serious sampling problems in many national data sets. Our article on average IQ by country explains why such rankings should be treated with great caution.
Why the Flynn effect matters
The Flynn effect is one of the clearest demonstrations that environment can shift measured intelligence for whole populations, and quickly. It also shows the limits of IQ as a fixed measuring stick: a score is always relative to a particular group at a particular time. The recent reversals are a reminder that those gains were not guaranteed, and that the environmental conditions behind them can change in either direction.