What is the Flynn effect?

Short answer: The Flynn effect is the rise in IQ test scores over time, which makes a test's norms go out of date (Trahan and colleagues, Psychological Bulletin, 2014). It is named after James Flynn, who described it systematically in 1984 and 1987. A 1995 American Psychological Association task force put the average gain at about three IQ points per decade. Later studies give smaller figures for recent decades, and some study populations, such as Norwegian male conscripts born after 1975, show a decline (Bratsberg and Rogeberg, PNAS, 2018). Nobody has settled why scores rose.

Key findings on the Flynn effect, by year

• 1984, James Flynn: every Stanford-Binet and Wechsler standardization sample from 1932 to 1978 set a higher standard than the one before. Flynn put the total rise in mean IQ for Americans at 13.8 points over 46 years (Psychological Bulletin).
• 1987, James Flynn: data from 14 nations showed gains of 5 to 25 points in a single generation, with some of the largest on culturally reduced tests and tests of fluid intelligence (Psychological Bulletin).
• 1995, an APA task force chaired by Ulric Neisser: its report says "The average gain is about three IQ points per decade; more than a full standard deviation since, say, 1940." The cause, the report said, "is presently unknown".
• 2014, Trahan and colleagues: a meta-analysis of 285 studies found a mean of 2.31 points per decade since 1951, and 2.93 points per decade for modern Stanford-Binet and Wechsler tests (Psychological Bulletin).
• 2015, Pietschnig and Voracek: a meta-analysis of 271 samples with almost 4 million people, covering 1909 to 2013, found the largest gains on fluid tests and found that gains "have decreased in more recent decades" (Perspectives on Psychological Science).
• 2018, Bratsberg and Rogeberg: among Norwegian male conscripts born 1962 to 1991, the rise, its turning point and the later decline all appeared within families, which points to environmental causes (PNAS).
• 2023, Dworak, Revelle and Condon: among 394,378 US adults tested online from 2006 to 2018, composite scores, matrix reasoning and letter and number series showed a reverse Flynn effect and verbal reasoning changed too little to count. 3D rotation, measured only in people tested from 2011 to 2018, showed a Flynn effect (Intelligence).
• 2024, Winter, Trudel and Kaufman: two studies in the WAIS-5 manual found a Flynn effect of only 1.2 IQ points per decade (Journal of Intelligence).

What the Flynn effect is

IQ scores are ranks. When a test is built, a large norm group takes it and its average is set to 100, and your score says where you stand against that group. Our guide to the average IQ explains how the 100 is set.

The Flynn effect is what happens to those norms over time. "The Flynn effect refers to the observed rise in IQ scores over time, which results in norms obsolescence," write Trahan, Stuebing, Fletcher and Hiscock in Psychological Bulletin (2014). People taking an old test today get more items right than its norm group did, so they score above 100 on the old norms.

Flynn was not the first to see rising scores. The APA task force report of 1995 says many psychometricians had noted the gains, and that Flynn (1984, 1987) "first described them systematically." His 1984 paper showed the pattern across every Stanford-Binet and Wechsler standardization sample from 1932 to 1978, and argued that obsolete norms had distorted hundreds of studies. Our guide to the Stanford-Binet test covers the oldest of those tests.

How big the rise was

The headline figure is about 3 points per decade. The APA task force gave it in 1995, and it matches Flynn's 1984 data: 13.8 points over 46 years is about 3 points per decade, by our calculation.

The estimates spread out once you look at more data:

• 2.31 points per decade across 285 studies since 1951, with a 95% confidence interval of 1.99 to 2.64 (Trahan and colleagues, 2014).
• 2.93 points per decade for 53 comparisons on modern Stanford-Binet and Wechsler tests since 1972 (Trahan and colleagues, 2014).
• About two to four points per decade since the early twentieth century, as Andrzejewski and colleagues summarize the research in a 2024 paper (Journal of Intelligence).
• Three to five points per decade from 1932 onward, as Northwestern University's 2023 news release puts it.
• 1.2 points per decade in the two WAIS-5 manual studies, against 2.6 to 2.8 in comparable WAIS-III and WAIS-IV studies more than 15 years earlier (Winter, Trudel and Kaufman, 2024).

These numbers differ because the samples, tests and periods differ. Read "3 points per decade" as a summary of the twentieth century on the major clinical tests.

Researchers also disagree on whether the effect is slowing. Trahan and colleagues found in 2014 that their modern-test estimate "was not consistent with the hypothesis that the Flynn effect is diminishing." A year later, Pietschnig and Voracek found that gains "have decreased in more recent decades" (2015). The WAIS-5 studies reported in 2024 found 1.2 points per decade, though they were small, with 186 and 98 people, which the authors list as a limitation.

Which abilities rose most

The gains were uneven. Pietschnig and Voracek estimated annual gains of 0.41 points for fluid IQ, 0.30 for spatial, 0.28 for full-scale and 0.21 for crystallized IQ (2015). By our calculation that is about 4.1, 3.0, 2.8 and 2.1 points per decade.

Fluid intelligence is reasoning on new problems, the kind a matrix puzzle asks for, as our guide to fluid intelligence explains. The APA task force noted in 1995 that "the largest gains appear on the types of tests that were specifically designed to be free of cultural influence", one of which is Raven's Progressive Matrices.

School subjects rose least. Andrade and Jamuna wrote in 2004 that gains "are least on subtests, such as arithmetic and information, that address school-taught subjects." The APA report adds a puzzle: SAT scores fell from the mid-1960s to the early 1980s while IQ scores kept rising (1995).

Why scores rose

Nobody knows for sure. The APA task force said in 1995 that the gains "seem much too large to result from simple increases in test sophistication," and added: "Their cause is presently unknown." It weighed more complex daily life and longer schooling, better nutrition, and Flynn's view that real intelligence cannot have risen that much.

Later work names similar candidates. A 2024 paper by Andrzejewski and colleagues lists "improved perinatal nutrition, access to medical services, and better schooling" as the most often cited (Journal of Intelligence). Some single explanations look weak. Andrade and Jamuna point out that gains were "recorded even before the television and video game era" (2004), and cite research that better diet gives only modest gains, clear only in children who were poorly nourished to begin with.

A study of Norwegian brothers narrowed the field. Bratsberg and Rogeberg used conscription scores for men born 1962 to 1991 and compared scores within families (PNAS, 2018). In their own plain-language summary, they write that the Norwegian Flynn effect and its reversal "were driven by environmental factors" and that "Any genetic effects present are negligible across our data period" (TheScienceBreaker, 2019). Their analysis "does not, however, speak to what these underlying environmental causes are."

The reverse Flynn effect

A reverse Flynn effect is a decline in scores across later cohorts within one population. Each finding below covers only the people in that study.

Norwegian male conscripts. The average score of Norwegian male conscripts "began to decline with the birth cohorts born after 1975" (Bratsberg and Rogeberg, TheScienceBreaker, 2019). Because the rise and the decline both appeared among brothers in the same family, the analysis "finds no evidence for prominent causal hypotheses of the decline implicating genes and environmental factors that vary between, but not within, families" (PNAS, 2018).

Men assessed for Danish military service. Teasdale and Owen studied this group. As the British Psychological Society's Research Digest reported in 2008, men tested in the late 1990s scored about 2 points better than those tested in 1988, and men tested in 2003/2004 scored close to the 1988 level again. We could read this study only through that secondary report.

US adults tested online. Dworak, Revelle and Condon analyzed 394,378 adults who answered ability items online from 2006 to 2018 (Intelligence, 2023). Composite scores, matrix reasoning and letter and number series showed "a pattern consistent with a reversed Flynn effect" across age, education and gender. According to the paper's abstract, verbal reasoning changed too little to meet the authors' threshold, and 3D rotation scores rose. The 3D rotation items were given only to people tested from 2011 to 2018, a subset of 303,540, according to Northwestern's news release. The release describes verbal reasoning as falling, but the abstract is the better guide.

Dworak cautioned against reading this as a fall in ability. In Northwestern's release she calls the decline "a difference in scores that are favoring older or newer samples," and adds that "not every domain is going down; one of them is going up." The release also notes that the study did not examine the reason, and that people may have paid less attention to ability items in a survey advertised as a personality test.

Not every recent study shows a decline. In standardization data from 1,392 Austrian residents tested between 2005 and 2018, Andrzejewski and colleagues found "positive Flynn effects across all domains of the IQ test" (Journal of Intelligence, 2024).

Platt and colleagues compared fluid intelligence test scores on norms from 1989 and 2003 in a representative sample of 10,073 US adolescents aged 13 to 18 (Intelligence, 2019). Overall, the Flynn effect in that sample was not significant. In the authors' adjusted estimates per decade, IQs rose 2.3 points at age 13 and fell 1.6 points at age 18 (Platt and colleagues, 2019).

Are people getting smarter?

Flynn himself thought not, at least not as much as the numbers suggest. The APA task force reported in 1995 that "Flynn himself believes that real intelligence" cannot have increased as much as the data suggest, and it closed that discussion with: "The issue remains unresolved." TED's page for his 2013 talk asks, "Are we actually getting smarter, or just thinking differently?" In the talk, Flynn's answer is that modern life trained new habits of mind, such as classifying things and taking hypothetical questions seriously.

A statistical test points the same way. Wicherts and colleagues checked whether five intelligence tests measured the same abilities across cohorts, using Dutch data such as Wechsler adult test scores from 1967 to 1999 plus reanalyses of earlier published studies (Intelligence, 2004). Their result: "measurement invariance with respect to cohorts is untenable." The tests did not measure exactly the same abilities in each cohort, so the rise in scores cannot be read as a simple rise in general ability.

What the Flynn effect means for your IQ score

For you, the Flynn effect is about the age of a test's norms. In a 2024 paper in the Journal of Intelligence, Winter, Trudel and Kaufman write that clinicians "became well aware that IQ test norms become out of date at the rate of three IQ points per decade" and that "Older norms tend to produce spuriously high IQs".

The Flynn effect is one reason publishers re-norm their tests. Winter and colleagues write that tests are re-normed traditionally every 10 to 15 years, and that the urgency to re-standardize grew "when the Flynn effect became an axiomatic fact of life, sometime in the 1990s" (2024). The Wechsler tests show the cycle (data years from Winter and colleagues, 2024; publication years from Pearson's product pages, read 18 October 2026):

• [WAIS-IV](/blog/wais-iq-test): norm data collected 2007 to 2008, published 2008.
• WISC-V: norm data collected 2013, published 2014.
• WAIS-5: standardized from February 2023 to January 2024 on 2,020 people aged 16 to 90, published 2024.

When a test is re-normed, the average is set back to 100. The APA task force gave an example in 1995: "if twenty years have passed since the last time the test was standardized, people who now score 100 on the new version would probably average about 106 on the old one."

Here is an illustration. Alex is invented for this example and scores 110 on a test whose norms are 15 years old. At 3 points per decade, the norms are about 4.5 points easy, so by our calculation the same performance would score about 105 to 106 on current norms. At the 1.2 points per decade found in the WAIS-5 studies (2024), the gap is about 1.8 points, or about 108 by our calculation.

Two cautions apply. These are population averages, not corrections to apply to your own score, and Andrade and Jamuna advise test users to "check the current validity of the norms that they are employing" (2004). And any single score already has a range: it can differ between tests and between sittings. Our guides to IQ percentiles and to whether IQ tests are accurate explain how to read that range.

If you compare an online score with a clinical one, ask when and on whom each test was normed. If you want to raise a score, our guide on whether you can increase your IQ covers what studies found.

Taking a test yourself

WhatIQ's test has 24 timed picture puzzles and takes about 15 minutes (whatiq.com, 18 October 2026). It gives a score on the standard IQ scale with the range it most likely sits in, and the site says the result is an estimate, not a clinical assessment.

Frequently asked questions

Who discovered the Flynn effect?

James Flynn, a New Zealand political scientist, described the Flynn effect systematically in two papers in Psychological Bulletin, in 1984 and 1987, and the effect carries his name. He was not the only one to see the rise. The APA task force report of 1995 says many psychometricians had noted the gains, and that Flynn "first described them systematically."

What is the reverse Flynn effect?

It is a decline in IQ test scores across later cohorts within one population. Among Norwegian male conscripts, the average score began to fall with men born after 1975, and the pattern appeared within families, which points to environmental causes (Bratsberg and Rogeberg, PNAS, 2018). In a 2023 study of US adults tested online from 2006 to 2018, composite, matrix reasoning and letter and number series scores fell, and 3D rotation scores, measured from 2011, rose (Dworak, Revelle and Condon).

Is IQ going down?

In some studies, yes, for the people those studies tested. Among Norwegian male conscripts, scores declined for men born after 1975 (Bratsberg and Rogeberg, 2018). Among US adults tested online from 2006 to 2018, composite, matrix reasoning and letter and number series scores declined (Dworak and colleagues, 2023). In standardization data from 1,392 Austrian residents tested between 2005 and 2018, scores rose in all domains (Andrzejewski and colleagues, 2024). No study we read measured "Gen Z" as a named group.

What causes the Flynn effect?

Nobody knows for sure. The APA task force wrote in 1995 that the cause of the gains "is presently unknown." The most often cited candidates are better nutrition around birth, access to medical services and better schooling (Andrzejewski and colleagues, 2024). Flynn pointed to modern life training more abstract habits of mind. A Norwegian within-family study found environmental causes, without naming which ones (Bratsberg and Rogeberg, 2018).

Does the Flynn effect affect my IQ score?

It can, through the age of the test's norms. Older norms tend to give scores that read high, at a rate long put at about three points per decade (Winter, Trudel and Kaufman, 2024). The WAIS-5 manual studies found only 1.2 points per decade, so the size depends on the test. Your score also has a range and can differ between tests and sittings.

How often are IQ tests re-normed?

Traditionally every 10 to 15 years, according to a 2024 paper by Winter, Trudel and Kaufman in the Journal of Intelligence. The WAIS-5 came 16 years after the WAIS-IV, which the authors say was longer than expected because of the COVID-19 pandemic. Each new norm sample resets the average to 100, so the rise between editions "more or less disappears from view", as the APA task force put it in 1995.

Is the Flynn effect still happening?

It depends on the study. Trahan and colleagues found in 2014 that modern-test data did not show it diminishing, while Pietschnig and Voracek found in 2015 that gains had decreased in recent decades. The WAIS-5 studies found 1.2 points per decade in 2024. Among Norwegian male conscripts, scores fell for men born after 1975 (Bratsberg and Rogeberg, 2018). In standardization data from 1,392 Austrian residents tested between 2005 and 2018, scores rose in all domains (Andrzejewski and colleagues, 2024).