In several wealthy, industrialized nations, the long-running rise in IQ scores known as the Flynn Effect does appear to have stalled or reversed, particularly among cohorts born after the late 1970s. But the picture is far from uniform. Developing and middle-income countries still show strong gains, and even within countries experiencing declines, the trend does not affect all ability levels or cognitive skills equally. The reversal is real in certain populations, but calling it a global phenomenon overstates the evidence.
What the Flynn Effect Actually Showed
For most of the twentieth century, each generation scored higher on standardized IQ tests than the one before it. A meta-analysis covering 285 studies and more than 14,000 test-administration pairs found the average gain was about 2.3 standard-score points per decade.1PubMed Central. The Flynn effect: a meta-analysis A broader meta-analysis spanning 1909 to 2013 put it slightly higher, around 2.8 points per decade, and found that roughly 90 percent of the annual changes across all studies were in the positive direction.2Perspectives on Psychological Science. One Century of Global IQ Gains: A Formal Meta-Analysis of the Flynn Effect (1909–2013) That amounts to a gain of about two full standard deviations over the century, which is enormous. A person scoring average in 1910 would, by the norms of 2010, look as though they scored well below average.
These gains showed up across countries, test types, and age groups. They were especially pronounced on tests of fluid reasoning, the kind that measure abstract problem-solving rather than learned facts. The gains were so reliable that test publishers had to re-norm their instruments every decade or so to keep the average at 100. The Flynn Effect became one of the most robust findings in psychology.
Where the Reversal Has Been Documented
The first clear signs of a slowdown came from Scandinavia, where military conscription data gave researchers unusually clean, large-scale records. In Norway, IQ scores among conscripts rose steeply from the mid-1950s through the late 1960s, gaining nearly 0.6 points a year. That pace slowed in the 1970s, and by the early 1980s scores had begun to fall.3Intelligence. The end of the Flynn effect?: A study of secular trends in mean intelligence test scores of Norwegian conscripts during half a century Denmark told a similar story: mean IQ rose from 100 for men born in 1940 to about 109 for those born in 1980, then started declining.4PubMed Central. The secular trend of intelligence test scores: The Danish experience for young men born between 1940 and 2000 That decline continued through at least the late 2000s in further Danish analyses.5Intelligence. The secular trend of intelligence test scores in the present century: The Danish experience
A systematic review found reports of a negative Flynn Effect in seven countries, mostly in Northern and Western Europe.6Intelligence. The negative Flynn Effect: A systematic literature review These included Norway, Denmark, the United Kingdom, Finland, France, the Netherlands, and Australia, though the magnitude and timing varied from country to country. The declines were modest, typically on the order of a few points per decade, but they were consistent enough to be taken seriously. A recent study focusing on the United States and several European nations argued that the reversal actually began earlier than previously assumed among sons of higher-income fathers, with lower-income groups peaking later because expanding educational attainment temporarily masked the underlying decline.7Proceedings of the National Academy of Sciences. Educational expansion and the hidden reversal of the Flynn effect—Differential trends across socioeconomic strata
The Developing World Is Still Gaining
While wealthy nations plateau or dip, the Flynn Effect appears alive and well in lower- and middle-income countries. A meta-analysis of studies using Raven’s Progressive Matrices, one of the most widely used nonverbal intelligence tests, found an overall gain of about 0.22 IQ points per year but noted that the strongest gains appeared in middle-income countries and among younger age groups.8Intelligence. Seventy years, 1000 samples, and 300,000 SPM scores: A new meta-analysis of Flynn effect patterns A review of the most recent data confirmed that pattern: positive Flynn Effects in economically less developed countries, but trivial and often negative effects in the most advanced ones.9Intelligence. Ongoing trends of human intelligence
Some of the clearest evidence for continuing gains comes from unexpected places. A study in rural Embu, Kenya, documented strong IQ gains between 1984 and 1998, with the biggest jumps on Raven’s matrices, the measure most sensitive to abstract reasoning.10Psychological Science. IQ on the Rise The explanation is intuitive: the same factors widely credited for driving gains in the West during the twentieth century, including better nutrition, more schooling, reduced infectious disease, and more cognitively stimulating environments, are still improving rapidly in much of Africa, South Asia, and Latin America. Once those improvements plateau, the gains likely will too.
Why the Environment Matters More Than Genetics
When scores start falling, one instinct is to wonder whether something genetic is happening. Maybe people with lower cognitive ability are having more children, slowly dragging down the population average over generations. This idea, sometimes called dysgenic fertility, has some empirical support as a measurable phenomenon. A study using a large Chinese dataset found a negative relationship between fertility and cognitive ability, estimating a loss of about 0.75 IQ points between 1971 and 2000 from this selection pressure.11Intelligence. Evidence of dysgenic fertility in China But the size of that effect is too small and too slow to account for the relatively rapid reversals seen in Scandinavian data.
The strongest evidence against a genetic explanation comes from within-family studies in Norway. Using military conscription records linked to administrative data, researchers compared brothers from the same family born in different years. If the reversal were genetic, brothers in the same family should score about the same, since they share their family’s genetic endowment. Instead, the researchers found that the entire Flynn Effect, its peak, and its decline could be recovered entirely from within-family variation. Brothers born later in the declining period scored lower than their older brothers, which means the cause has to be something that changed in the environment between their births, not something about which families were having children.12PubMed Central. Flynn effect and its reversal are both environmentally caused Earlier work on Norwegian brothers’ data reached a similar conclusion, showing that the intelligence differences between siblings tracked closely with the direction of the Flynn Effect during the period they were born.13Intelligence. The Flynn effect in sibships: Investigating the role of age differences between siblings
A meta-analytic investigation likewise found little evidence that migration, fertility patterns, or mortality could substantively explain IQ decreases.14Politische Psychologie. Is the Flynn effect related to migration? Meta-analytic evidence for correlates of stagnation and reversal of generational IQ test score changes This does not mean demographic factors are irrelevant, but they do not appear to be the primary driver. Whatever is causing the reversal, it is something in the shared environment that shifts quickly enough to show up between siblings born just a few years apart.
Candidate Explanations for the Decline
If the reversal is environmental, the obvious next question is which environmental factor changed. No single explanation has won consensus, but a few candidates keep surfacing.
Education is the most frequently cited. The original Flynn Effect is widely attributed, at least in part, to expanding access to formal schooling. Once nearly everyone in a country finishes secondary school, that particular tailwind fades. A British study noted that a recent IQ decline coincided with a drop in the proportion of students entering three-year advanced-level programs for 16-to-18-year-olds.15Personality and Individual Differences. A long-term rise and recent decline in intelligence test performance: The Flynn Effect in reverse The U.S.-European analysis mentioned earlier formalized this idea, arguing that educational expansion temporarily masked an underlying decline, and once it peaked, scores fell across all income groups.7Proceedings of the National Academy of Sciences. Educational expansion and the hidden reversal of the Flynn effect—Differential trends across socioeconomic strata The implication is sobering: the educational system may have been propping up scores for decades, and what we are seeing now is what happens when that support plateaus.
Changes in how people spend their time, particularly the rise of digital media, are another popular suspect. A review of the research on digital device usage and childhood cognitive development noted evidence linking screen time to negative effects on brain volume and intelligence scores in some studies.16PubMed Central. Digital Device Usage and Childhood Cognitive Development: Exploring Effects on Cognitive Abilities But the evidence here is messy and contradictory. A large longitudinal study of U.S. children found that after controlling for genetics and socioeconomic background, time spent watching videos and playing video games was actually associated with small IQ gains over two years, not losses. Social media screen time, by contrast, showed no significant independent effect in either direction.17Scientific Reports. The impact of digital media on children’s intelligence while controlling for genetic differences in cognition and socioeconomic background Blaming screens for the reversal is popular in opinion pieces, but the controlled evidence so far is too mixed to call it a confirmed cause.
Other proposed factors include changes in nutrition quality (different from nutrition quantity, which improved throughout the twentieth century), declining reading habits, shifts in pedagogical methods, and reduced exposure to cognitively complex work. None of these has been isolated as the definitive cause. It is plausible that no single factor is responsible, and that the reversal reflects a cluster of environmental changes that each contribute a small amount in the same direction.
Not All Abilities Are Moving in the Same Direction
One reason the reversal is hard to pin down is that it does not hit all cognitive skills equally. The original Flynn Effect was strongest on measures of fluid intelligence, which capture abstract reasoning and pattern recognition. But the reversal pattern varies by subtest and by country. Data from German-speaking countries show verbal gains alongside spatial losses among adults, while Nordic data covering a range of subtests reveal a more mixed picture across vocabulary, analogies, number series, and spatial reasoning.18Intelligence. IQ decline and Piaget: Does the rot start at the top?
This unevenness matters because it constrains which explanations are plausible. If the decline were simply about worsening education, you would expect crystallized abilities like vocabulary and general knowledge to drop. If it were about reduced engagement with abstract problem-solving, fluid reasoning should take the biggest hit. The mixed pattern suggests that whatever is happening is not reducible to a single mechanism acting on a single type of thinking.
The Reversal Is Not Uniform Across Ability Levels
Perhaps the most striking wrinkle in the data is that the reversal does not affect everyone the same way. A population-based study of about 10,000 U.S. adolescents found that fluid IQ scores dropped nearly 5 points for those at the lower end of the distribution (IQ at or below 70) while rising about 3.5 points for those at the upper end (IQ at or above 130).19Intelligence. The Flynn effect for fluid IQ may not generalize to all ages or ability levels: A population-based study of 10,000 US adolescents In other words, the highest-scoring adolescents were still gaining while the lowest-scoring ones were losing ground, and the population average reflected a tug-of-war between these opposing trends.
This finding complicates the narrative that “IQ is falling” in a blanket sense. It suggests that the forces driving the reversal may disproportionately affect people who were already at a cognitive disadvantage, potentially because those at the top of the distribution have access to environments that continue to support cognitive development even as average conditions worsen. The socioeconomic stratification identified in the U.S.-European study fits this pattern: the decline began earliest among higher-income families (who had already exhausted the gains from educational expansion) but became steepest in that group after the educational plateau, while lower-income groups continued gaining for longer thanks to ongoing educational access improvements.
Could the Tests Themselves Be the Problem?
Before concluding that people are actually getting smarter or dumber, it is worth asking whether the measuring instruments are behaving consistently. IQ tests are periodically updated and re-normed, and comparisons across editions assume that the tests are measuring the same thing. That assumption does not always hold. An analysis of the Wechsler Adult Intelligence Scale found that scores were equivalent between the third and fourth editions, but not between the second and third editions, meaning that some of the published Flynn Effect findings may reflect changes in the instrument rather than genuine shifts in cognitive ability.20PubMed. Using Score Equating and Measurement Invariance to Examine the Flynn Effect in the Wechsler Adult Intelligence Scale
This is not a reason to dismiss the Flynn Effect or its reversal outright. The Scandinavian conscription studies used the same test across decades, sidestepping the instrument-change problem, and they still found both the rise and the decline. But for countries where the evidence comes from comparing different test editions, the measurement issue adds real uncertainty. Any individual study claiming a gain or loss of a few IQ points could be partly or entirely an artifact of the test changing rather than the population changing.
What Experts Expect Going Forward
A survey of 75 intelligence researchers asked them to predict IQ trends for the twenty-first century. Their consensus: substantial gains of six to seven points in regions where average scores are currently low, including Latin America, Africa, India, and East Asia, but essentially flat trends in the West, with a small predicted decline of about half a point in the United States.21Personality and Individual Differences. Survey of expert opinion on intelligence: The FLynn effect and the future of intelligence The experts attributed the original twentieth-century rise primarily to better health, nutrition, and education, and they pointed to the plateau of those same factors as the likely reason for stagnation in wealthy countries. When asked about causes of a possible reversal, the experts cited differential fertility, changes in educational quality, and media influences.
The expert forecast fits the data reasonably well. It portrays a world where the Flynn Effect is not so much ending as migrating: fading in countries where its drivers have been exhausted, but continuing in places where improvements in living conditions are still accelerating.
Growing Brains and Cognitive Reserve
An interesting parallel to the IQ trend comes from brain imaging. A study of participants in the Framingham Heart Study, spanning birth cohorts from the 1930s through the 1970s, found that people born more recently had larger intracranial and cerebral volumes, including larger hippocampi and more cortical gray matter, even after adjusting for body size.22PubMed Central. Secular Trends in Head Size and Cerebral Volumes In the Framingham Heart Study for Birth Years 1902–1985 Later-born cohorts also performed better on memory tests. The researchers suggested that these increases in brain volume could contribute to greater cognitive resilience, potentially offering some protection against age-related cognitive decline and diseases like Alzheimer’s.23PubMed Central. Trends in Intracranial and Cerebral Volumes of Framingham Heart Study Participants Born 1930 to 1970
Whether these brain-volume trends will continue, plateau, or reverse in cohorts born after the 1970s is not yet known. If the environmental factors that drove bigger brains overlap with those that drove higher IQ scores, the same deceleration could eventually show up in neuroimaging data. But for now, the brain-volume finding is a reminder that the Flynn Effect had a physical substrate: it was not just about test-taking habits or familiarity with test formats. Something genuinely changed in how human brains developed over the twentieth century.
Why This Has Economic Stakes
IQ trends are not just an academic curiosity. Population-level cognitive ability is correlated with economic productivity, and even small shifts matter when compounded over millions of people and decades of time. One modeling study estimated that each IQ point corresponds to about $810 in higher average per-capita productivity per year. Based on projected trends among 17-year-olds in the United States, the researchers estimated a productivity gain of roughly $1,750 to $3,000 per person between 2012 and 2060, depending on which model was used.24PLOS ONE. Future Cognitive Ability: US IQ Prediction until 2060 Based on NAEP Those estimates assumed continued modest gains. If scores instead stagnate or decline, the economic implications could run in the opposite direction.
This kind of modeling carries substantial uncertainty, and the dollar-per-IQ-point figure should be treated as a rough approximation. But the general point holds: if the cognitive gains of the twentieth century underpinned some portion of economic growth, a reversal of those gains is not just a question for psychometricians. It touches workforce readiness, innovation capacity, and the economic trajectories of entire countries.
How Different Subtests Respond to Cultural Change
One underappreciated aspect of the Flynn Effect, and its possible reversal, is that “IQ” is not a single thing being measured. A full IQ battery includes subtests for vocabulary, working memory, processing speed, spatial reasoning, and abstract logic, among others. These subtests have historically moved at different rates. The original Flynn Effect was largest on fluid measures like Raven’s Progressive Matrices and smallest on crystallized measures like vocabulary or general knowledge. This made sense: the environmental improvements of the twentieth century, such as more schooling, richer visual environments, and more practice with abstract categorization, would naturally boost the type of thinking those fluid tests tap into.
The reversal, where it has occurred, does not mirror the original gain in reverse. It is not simply the case that the subtests that gained the most are now losing the most. The mixed pattern observed across countries, with some populations showing verbal gains alongside spatial losses, suggests that modern environments may be reshaping the cognitive profile of the population in ways that resist a single narrative. People may be getting better at some types of thinking while getting worse at others, and the net effect depends on which subtests a given IQ battery emphasizes. A test that leans heavily on spatial reasoning will tell a different story about the reversal than one that emphasizes verbal comprehension.
This matters practically because the choice of test can determine whether a country appears to be in decline or still gaining. It also matters theoretically, because if the reversal is driven by specific environmental changes, those changes should leave fingerprints on specific cognitive skills rather than dragging everything down uniformly. Tracking which abilities are falling and which are holding steady, or even improving, is probably the best route to identifying the cause.