The Flynn effect: why IQ scores rose
The Flynn effect is the steady rise in average scores on intelligence tests during the twentieth century, by roughly three IQ points per decade in many countries. It is named after James Flynn, the researcher who documented how large and widespread it was. Nobody knows for certain what caused it. The main candidates are more and better schooling, improved nutrition and health, and daily life that demands more abstract thinking. In several countries the rise has slowed or reversed since the 1990s.
What James Flynn documented
James Flynn was a researcher at the University of Otago in New Zealand. In the 1980s he gathered data showing that each new generation did better on intelligence tests than the one before, first in the United States and then in more than a dozen other countries. Others had noticed rising scores earlier, but Flynn showed that the pattern was large, consistent and international, and he made it impossible to ignore.
The effect is normally invisible because of how IQ scores work. A score does not count correct answers. It compares you with a reference group, and the average of that group is defined as 100 (see what IQ is). When a test is updated, a new reference group sets the new average. Flynn looked at what happened when the same people took both an old and a new version of a test. They regularly scored higher on the old version, because they were being compared with an earlier generation that had got fewer items right.
How large the effect was
The usual summary is about three points per decade, or around thirty points over a century. That is two standard deviations on a scale where one standard deviation is 15. The exact rate differed between countries, periods and types of test, so the figure is a rough average and not a law.
Taken at face value, it leads to an odd conclusion. Scored against today's norms, the average person of a hundred years ago would land at around 70. Flynn himself did not believe that our great-grandparents were barely able to cope with life. His reading was that the tests measure particular mental habits, and that those habits had changed.
One detail supports this. The gains were not evenly spread. They were largest on tests of abstract reasoning, such as Raven's matrices and tasks that ask what two things have in common. They were much smaller on vocabulary, arithmetic and general knowledge. The difference between these kinds of ability is explained in fluid vs crystallized intelligence.
The proposed explanations
The rise was too fast to be genetic, since the genetic make-up of a population does not change that much within a few generations. The causes must lie in the environment. Which ones, and in what proportion, is still debated. These are hypotheses, and most researchers think several of them worked together.
| Hypothesis | The idea |
|---|---|
| Schooling | People stayed in school far longer, and teaching moved from memorising towards classifying, reasoning and working with symbols. |
| Nutrition and health | Better food, fewer childhood infections and better medical care supported physical and brain development, much as average height increased. |
| Abstract thinking in daily life | Work, media and technology became more complex. People got used to thinking in categories and hypothetical cases. This was Flynn's preferred explanation. |
| Smaller families | With fewer children per family, each child may receive more adult attention and resources. |
| Familiarity with tests | Multiple-choice formats, timed tasks and puzzles became a normal part of life, so tests felt less strange. |
Each explanation has weak points. Test familiarity, for example, is generally thought to account for only a small part of the rise. Schooling has the most direct support, since additional years of education are linked to higher scores, as discussed in can you increase your IQ.
Did people really become more intelligent
That depends on what you mean. People clearly got better at the specific kinds of thinking the tests ask for. Flynn described it as putting on “scientific spectacles”: learning to sort the world into abstract categories and to reason about hypothetical situations, instead of thinking mainly in terms of concrete, practical use.
That is a real change in how people think. It is not the same as saying that people a century ago were less capable in general. Their thinking was fitted to different demands. Most researchers therefore treat the Flynn effect as a change in particular skills, not as a thirty-point rise in general intelligence.
The recent levelling off and reversal
The rise has not continued everywhere. Since the 1990s, several countries with good long-term data have recorded a levelling off or a decline. Norway, Denmark and Finland are the best known examples, because they tested large groups of young men at military conscription year after year. This is sometimes called the negative or reverse Flynn effect.
The reasons are as uncertain as the reasons for the rise. One Norwegian study found that the decline also shows up within families, with younger brothers scoring lower than older brothers. That points to environmental causes and not to a change in who is having children. Suggested factors include changes in education and in how people spend their time, and the possibility that the earlier gains had simply reached their limit. None of these is established.
The picture is not the same worldwide, and in some countries scores are still rising. Because data quality differs so much from place to place, comparisons between countries are unreliable.
What it means for reading a score
The practical consequence is that an IQ score is only meaningful against current norms. Test publishers re-norm their tests every so often, testing a new representative sample and resetting the average to 100. Without that, scores would drift upward and stop meaning what they are supposed to mean.
- Old tests give inflated scores. If a test is scored against norms that are decades old, you are compared with an earlier generation, and your result will usually come out higher than on a current test.
- A score is relative. 100 always means the average of the reference group at the time the test was normed. It is not a fixed amount of intelligence. The IQ scale describes positions within that group.
The Flynn effect is also a reminder to hold any single number loosely. If the average of whole populations can shift by many points within a few generations, test scores clearly respond to the conditions people live in. An online test, ours included, adds further uncertainty and gives an estimate, not a clinical result.
Keep reading
What is IQ?
IQ is a score that compares your result on a reasoning test with people of your age. What the number means, where it comes from and what it leaves out.
What is the average IQ?
The average IQ is 100 because tests are built that way. How the bell curve spreads scores around it, and what share of people falls in each range.
IQ scale: what each score range means
The usual IQ score ranges on a scale with mean 100 and standard deviation 15, how many people fall in each, and why a score is a range, not a point.