In the 1980s, New Zealand political scientist James Flynn noticed a strange pattern when comparing IQ test norms from different years. Test publishers re-normed their tests every 15 to 20 years, and the same thing happened every time: when scored against older norms, the new generation scored well above average.
The increase was roughly 3 IQ points per decade. It continued throughout the twentieth century, appeared in dozens of countries, and is known today as the Flynn effect.
What exactly is the Flynn effect?
The Flynn effect is the systematic increase in raw test performance across generations. It stays hidden because tests are constantly re-normed: since the average is always set to 100, the rise does not show up in final scores.
To see the effect, you have to give an older test to a modern group. When you do, an average adult today scored against 1950s norms lands in the 115–120 range. By that era's standards, that is "well above average."
Flynn effect summary table
Gains measured in different countries and periods look roughly like this:
| Region | Period | Decade gain (approximate) |
|---|---|---|
| US | 1932 – 1978 | 3 points |
| Western Europe | 1950 – 1990 | 3 – 5 points |
| Japan | 1940s – 1960s | Near 7 points in some periods |
| Developing countries | Post-1970 | Rise continues |
The exact numbers vary across studies, but the direction stays the same. The only exception is the last thirty years, which we look at below.
The increase is not equal across all areas
This is the most overlooked and revealing detail of the Flynn effect. Gains are not spread evenly across every section of the test:
- The largest gains are in abstract pattern tests. In sections like Raven's Progressive Matrices, which require no vocabulary or prior knowledge and rely purely on pattern recognition, gains reach up to 5 to 7 points per decade.
- The smallest gains are in vocabulary and knowledge sections. General knowledge, arithmetic, and vocabulary subtests show almost no gain.
This breakdown is crucial. If people were simply "smarter" overall, vocabulary and general knowledge would rise at the same speed. They do not. What rises is abstract reasoning, which measures fluid intelligence. To explore this distinction further, our article on fluid intelligence and crystallized intelligence covers the topic in detail.
Why scores are rising
There is no consensus on a single explanation. The leading hypotheses include:
Nutrition and health
During the twentieth century, childhood nutrition, infant mortality rates, and the burden of infectious diseases improved dramatically. Early childhood illness is known to limit cognitive development. This hypothesis explains why gains were concentrated at the lower end of the distribution.
Length and quality of education
Average schooling grew from a few years to well over ten years in a single century. School does not just provide knowledge; it also trains people to solve tests and follow abstract instructions.
Abstract thinking in everyday life
This is James Flynn's preferred explanation. He argued that modern life forces people to view the world through categories and hypotheticals. Flynn often cited a classic question asked to rural participants in the early 20th century: "What do a dog and a rabbit have in common?" The answer was usually "You hunt rabbits with dogs"—a concrete and functional response. Today, the expected answer is "They are both mammals"—an abstract and classification-based response. IQ tests reward this second type of thinking.
Smaller family sizes and test familiarity
Increased adult attention per child and general familiarity with testing may have also contributed. However, test familiarity alone fails to explain a steady increase lasting several decades.
The reverse Flynn effect: Has the rise stopped?
Starting in the mid-1990s, research from several Nordic countries began showing a reversal of the trend. Analyses of decades of conscription data from Norway showed that scores began dropping after the generation born around 1975. Similar findings emerged from Denmark and Finland.
What makes this notable is that the decline in the Norwegian data occurs between siblings in the same family. This largely rules out explanations like immigration or changing population demographics, pointing instead to environmental factors. Changes in educational curricula, declining reading habits, and shifting media consumption are among the suggested causes, though none are proven.
Were our ancestors really less intelligent?
Flynn's own answer was no. In his view, what increased was not raw mental capacity, but a mindset tailored to a specific way of thinking. A farmer a century ago was highly skilled at solving complex problems in their own context; they were simply unfamiliar with the abstract categorization language rewarded by IQ tests.
This serves as an important reminder of what an IQ score represents: it is not an absolute measure of intelligence, but a person's standing within their generation on a specific set of tasks. We discussed this in our post on how IQ scores are calculated.
Practical implications for testing
The Flynn effect has a concrete practical consequence: tests scored using outdated norms inflate results. If a test was normed 20 years ago, someone taking it today will score a few points higher than they should.
This is a serious issue in clinical evaluations, where it can alter decisions near diagnostic cutoffs. It is also why tests must be re-normed regularly. For many tests found on the internet, the date of the norming data is unclear—we discussed the limits of online tests in detail in this article.
Was the Flynn effect observed in Turkey?
Long-term, continuous norm datasets for Turkey are limited. However, rapid improvements in education duration and health indicators suggest a similar rise likely occurred. Comparable norming studies would be needed to know for certain.
Can the rise continue forever?
It did not. In developed countries, the curve flattened and, in some cases, reversed. This supports the idea that the effect is an environmental gain rather than biological evolution—once environmental conditions reach saturation, gains stop.
Which norm is used to calculate my score?
For professional tests, the norm year is listed in the manual. Online tests usually omit this information, so it is best to view your result as a rough estimate rather than an exact measure. If you want to see your score, you can take our free IQ test.