The Flynn effect is the long-run rise in raw IQ test scores: about 3 points per decade in industrialised countries across most of the twentieth century, named after James Flynn, who documented it in the 1980s. Nutrition, schooling, smaller families, less disease and a more abstract, symbol-heavy daily life are the main explanations. Since the 1990s the gains have slowed or reversed in several Western countries. Tests are renormed every 10–15 years to keep the average at 100, which is why old and new scores are not comparable.
What Flynn found
Test publishers renorm their tests periodically. When they do, they give both the old and the new version to the same sample. James Flynn noticed in 1984 that every time this was done in the United States, the sample scored higher on the old test than on the new one — about three points per decade of gap between the norming dates. He then found the same pattern in 14 countries, with gains ranging from 5 to 25 points per generation. Someone who scored 100 on a 1950 test would score around 80 on 2000 norms; someone scoring 100 today would have scored around 120 against 1950 norms.
The gains were largest on tests of fluid reasoning such as Raven's matrices and smallest on vocabulary and arithmetic, which is a clue to the cause.
The leading explanations
- Nutrition and health. Better food, fewer infections, iodised salt and less lead exposure all raise cognitive performance; the gains were steepest where childhood conditions improved fastest, and the effect shows up at the bottom of the distribution more than the top.
- Schooling. More years in education, and education that trains abstract classification (“how are a dog and a rabbit alike?”) rather than practical recall.
- Smaller families and more stimulating homes. Fewer children per family and more attention per child.
- “Scientific spectacles”. Flynn's own favourite: modern life asks people to think in hypotheticals, categories and symbols constantly — maps, screens, rules of games — and matrix tests reward exactly that.
- Test familiarity. People have become more used to multiple-choice formats and timed tasks.
Genetic change is ruled out by the speed: gene frequencies cannot shift that much in two generations. The Flynn effect is one of the clearest demonstrations that environment moves IQ, which is worth remembering when reading about heritability.
Has the Flynn effect reversed?
In Norway, Denmark, Finland, the Netherlands, France and the UK, cohorts born after about 1975 show flat or declining scores on the same military and school tests that had been rising for decades. The best-known study, Bratsberg and Rogeberg (2018), used Norwegian conscription records for more than 700,000 men and found the decline occurring within families — between brothers — which points to environmental causes (changes in schooling, media, nutrition patterns) rather than demographic ones.
The reversal is modest, a few points, and not universal; developing countries are still gaining. Whether it is a plateau, a measurement artefact or a real decline in the abilities the tests sample is still debated.
What it means for your score
Two practical consequences. First, an IQ score is only meaningful against the norms it was computed on, and those norms have a date. A score from a test normed in 1990 overstates your standing against today's population by several points, which matters for gifted programmes and clinical cut-offs. Second, comparing your score with a parent's or grandparent's score from decades ago is comparing against different yardsticks.
Online tests have the same issue in a different form. A test calibrated against a self-selected sample of people who choose to take IQ tests online is measuring you against that group, not against a national census. We say so plainly on the methodology page and in the data study, whose sample averaged 114.5 for exactly that reason.
Frequently asked questions
How much do IQ scores rise per decade?
About three points per decade on average in industrialised countries during the twentieth century, with large variation by country and by type of test.
Are people smarter than their grandparents?
On the tests, yes, by roughly 20–30 points over three generations. Whether this reflects greater general intelligence or a shift toward the abstract thinking the tests reward is exactly what Flynn spent his career arguing about.
Why is the average IQ always 100?
Because tests are renormed every 10–15 years so that the current population averages 100. The Flynn effect is visible only when the same people take tests with different norming dates.
Is the Flynn effect still happening?
In many developing countries, yes. In several Western European countries scores have been flat or slightly falling for cohorts born since the mid-1970s.
Measured against today
Take the free IQ test and read the methodology page to see exactly how your score is computed and against whom.
Take the Free IQ Test