If an adult today took an IQ test normed in 1950, they would probably score well above 100. Not because they are exceptional, but because the whole population has improved on this kind of test. The phenomenon has a name: the Flynn effect.
A statistician's discovery
In the 1980s, New Zealand political scientist James Flynn looked into a technical question: what happens to IQ scores when you compare old and new versions of the same test? In 1984, he showed that in the United States, scores had risen by about 3 points per decade between the 1930s and the late 1970s. In 1987, he extended the analysis to 14 countries: the rise was everywhere, sometimes dramatic.
Other researchers had noticed it before him (as early as 1948, psychologist Reuel Tuddenham observed that American recruits in World War II did far better than those of World War I). But Flynn measured its scale, and the phenomenon ended up bearing his name.
Huge gains, but not everywhere
The most striking example comes from the Netherlands, where 18-year-old men took a reasoning test during military service. Between 1952 and 1982, their average score rose by about 20 points: more than one standard deviation in a single generation.
The gains are not spread evenly across tasks:
- they are very large on abstract reasoning tests, such as Raven's matrices or "similarities" (finding what two ideas have in common);
- they are small on vocabulary, arithmetic and general knowledge.
In other words, it's not that we know much more than our grandparents: what has changed is mostly the way we reason about abstract problems.
Why did IQ go up?
Such a fast rise cannot be genetic: a population's genes don't change in thirty years. Researchers put forward several explanations, which probably work together:
- nutrition and health: over the same period, average height also increased, and childhood infectious diseases, which weigh on brain development, declined;
- schooling: people stay in school much longer, and school trains precisely the skill of handling categories and hypotheses;
- smaller families, with more attention and stimulation for each child;
- a more abstract environment: knowledge jobs, symbols, screens, maps, diagrams… modern life constantly asks us to reason about things we can't see.
Flynn stressed this last point. He talked about "scientific spectacles": we have grown used to sorting things into abstract categories. Ask someone today what a dog and a rabbit have in common, and they will say "they're both mammals". In the 1930s, Soviet psychologist Alexander Luria interviewed peasants in Central Asia with little schooling, who tended to answer in concrete, useful terms: you use a dog to hunt rabbits. A clever answer, but one that earns no points on an IQ test.
The paradox
Taken literally, the figures suggest that the average person in 1900 would score around 70 by today's norms. Yet our great-grandparents were nothing like intellectually disabled: they ran farms, shops and families.
For Flynn, this paradox shows that IQ doesn't measure some pure, unchanging "intelligence", but habits of thinking that depend on the era. Underlying abilities did not necessarily explode; the skills that tests reward became more widespread.
The reversal
Since the 1990s, several countries have seen the trend slow down or even reverse. The strongest study comes from Norway: in 2018, economists Bernt Bratsberg and Ole Rogeberg analysed decades of tests taken by Norwegian conscripts. Scores rose for men born up to the mid-1970s, then fell.
An important detail: the decline shows up within the same families, between older and younger brothers. So it isn't caused by a change in the population, but by the environment, although nobody yet knows exactly which part of it (changes in schooling, reading habits, screens, the tests themselves…). Denmark and Finland have reported similar trends.
What it changes in practice
The Flynn effect has a very practical consequence: tests age. If a test's norms are twenty years old, it overestimates scores, because the reference population has changed. That is why major tests like the WAIS are re-normed regularly.
In the United States, the issue has even reached the courts: since 2002, the death penalty cannot be applied to a person with an intellectual disability, and the date a test was normed has sometimes been debated at trial.
In short
- IQ scores rose by about 3 points per decade in the 20th century, mostly on abstract reasoning.
- The causes are environmental: health, nutrition, schooling, a more abstract world.
- In some countries the trend has reversed for generations born after the mid-1970s.
- IQ also measures habits of thinking tied to an era, not a fixed capacity.
To find out what an IQ test measures today, read our article What does an IQ test really measure?, then try the matrices in the Viberankly IQ test.