In the past decade, researchers have found data suggesting that, on average, the populations of developed Western countries are becoming more stupid. It’s a reversal of the so-called Flynn Effect, named after the New Zealand philosopher and intelligence researcher James Flynn. In the 1980s, Flynn had shown that, between the 1930s and the 1970s, average intelligence rose by three IQ points per decade in the United States and in other Western nations, a trend that seemed to persist until the 1990s. But then in the early years of the 21st century, Flynn noticed something disturbing: IQ scores for even the brightest children in the US and the UK had started to decline. And in Nordic nations, Flynn projected that average national intelligence scores – some of which had been declining since the mid-1990s – could drop by around seven IQ points over the following 30 years.
Plenty of reasons are given for this decline. In November 2025, Lane Brown, a feature writer for New York magazine, cited several possible factors: outsourcing our cognitive labour to AIs, being glued to our screens, the ongoing effects of COVID-19 and, though this is probably a niche factor, too-strong weed. ‘The world is dumber, and we all know it,’ writes Brown. ‘Lately, it feels like that culturewide upgrade to our mental operating systems has been rolled back to an older and buggier version.’
The negative Flynn Effect, however, isn’t what it seems. Firstly, falling average intelligence among Western nations is perfectly compatible with rising intelligence rates among certain sections of their populations. But more worryingly, the IQ tests on which Flynn relies are of doubtful value. Our stupidity, it turns out, is not always easy to understand.
A century ago, the superbly named pioneer of intelligence tests, Professor Edwin G Boring, remarked that ‘Intelligence is what the tests test.’ It’s a maxim whose circularity suggests that aptitude at passing IQ tests may be a hopeless proxy for intelligence and doesn’t indicate much about real-life cognitive abilities. Instead, IQ scores reward what society regards as valuable mental traits at a particular slice of time: your IQ number is a judgment rather than an objective fact. And yet, that number may serve as a curse or a badge of honour, depending on which side of the average 100-point score a person falls. As Brown puts it, ‘both the original Flynn effect and its reversal might owe more to inconsistent methodology than to real cognitive change.’
IQ tests can also be unhelpfully culturally specific, with disastrous, even racist results. In the BBC documentary Subnormal (2021), one of the UK’s first Black educational psychologists, Waveney Bushell, told an interviewer that underperforming Black kids who were segregated into special educational needs schools were often considered to be educationally subnormal due to unhelpful culturally specific questions on IQ papers. For instance, one question invited children to identify a ‘tap’ – more commonly known as a ‘faucet’ in the US. Easy enough, one would have thought. But some Black kids struggled with the question. It was not because they were stupid, but because in parts of the Caribbean, where their families are from, the word ‘pipe’ is used rather than ‘tap’.
For Arendt, our leading moral task is to overcome that stupidity – that form of moral blindness
Perhaps there’s more to intelligence than passing an IQ test. Consider the case of what Newsweek, in a 1945 article, called ‘the evil geniuses’. Before high-ranking Nazis were tried at Nuremberg, American psychologists measured their IQ scores and, worryingly, found that, by that metric, many of the most evil men in the world were geniuses: the economics minister Hjalmar Schacht scored 143, the Air Force commander Hermann Göring 138, and Hitler’s architect Albert Speer 128.
In 1961, when the philosopher Hannah Arendt looked across a Jerusalem courtroom into the eyes of Adolf Eichmann, one of the leading organisers of the Holocaust, she was struck by his absolute inability ‘to think from the standpoint of somebody else’. That is precisely what Arendt meant by the ‘banality of evil’. Then in an interview in 1964, Arendt said: ‘It was his thickheadedness that was so outrageous, as if speaking to a brick wall. And that was what I actually meant by banality … There’s simply resistance ever to imagine what another person is experiencing.’ For Arendt, our leading moral task is to overcome that stupidity – that form of moral blindness.
So, even if our intelligence measures really are flawed, what explains the rise and fall of IQ scores? Flynn hypothesised the importance of environmental factors, including changes in education, nutrition, family structure, economic pressures, microplastics, antidepressants and woodfire smoke. But that’s not an adequate explanation: from the 1930s, when average national intelligence rose in developed Western nations decade on decade, other deleterious environmental factors were at play. There was a lack of free public education and socialised healthcare, and toxic air pollution – a known factor influencing intelligence scores and cognition – was often unchecked. Complicating things further, what Flynn proposed as negative factors may be positive ones, and vice versa. Antidepressants, for instance, might help with improving one’s intelligence precisely by lifting one out of existential hopelessness.
More recently, observers have suggested two knockdown factors accounting for the apparent declines in average national intelligences: screens and social media. For many, these are intuitively seen as accelerants to the 21st century’s bonfire of stupidity – we seem to collectively sense that they are dumbing us down.
Intuitively, the argument that digital media is dumbing us down is plausible
On this point, it’s worth noting that the scores used to support the negative Flynn Effect weren’t down in every category. In 2023, researchers at Northwestern University and the University of Oregon published a study indicating that average American intelligence declined between 2006 and 2018 across three of four broad domains tested. They found that their fellow citizens tracked falling scores in logic, vocabulary, visual and mathematical problem-solving, and analogical reasoning. Only scores for spatial ability – the measure of the mind’s ability to analyse three-dimensional objects – rose during that period for the average American. Again, this highlights the culturally specific value judgment of those who contend that IQ tests demonstrate falling average intelligence. Perhaps spatial ability, for instance, is more important to today’s cognitive elites than reading. Certainly, you’ll need the former if you’re going to be any good at fast, visually intensive online games such as Fortnite. It is uncertain right now whether the gaming virtuosos who will play the next iteration of the Grand Theft Auto series will be dumber than someone who can write a plausible essay on the depiction of female autonomy in Samuel Richardson’s literary classic Clarissa; or, The History of a Young Lady (1748). Perhaps both skills are signs of intelligence; perhaps neither. What’s important here is that it’s hard to determine whether the latter is a more objective measure of intelligence than the former.
Intuitively, the argument that digital media is dumbing us down is plausible. And book after book, based on study after study, appears to confirm our intuitions. Computers and smartphones spare us cognitive labour. As a result, the human mind has less to do. Therefore, the need to be intelligent is less of an evolutionary imperative than it was in the pre-digital era. Computers and smartphones are more complex than ever, but human routines are oddly simpler. Generations ago, dishwashers and clothes dryers eased physical labours in daily life. Today, an iPhone or an Amazon Echo can ease mental labour, enabling us, the creators of these machines, to slide into the warm bath of mental fatuity. But perhaps, in principle at least, using an Echo or talking with ChatGPT allows you to free your mind for more cognitively challenging work.
The Canadian media theorist Marshall McLuhan saw the invention of the light bulb as an enlightening moment for humanity. ‘A light bulb creates an environment by its mere presence,’ he wrote in Understanding Media (1964) – the light extends human powers over what was hitherto dark. It is not utter folly to believe that AI and other digital tools derided for making us more stupid work in similar ways, by extending human cognitive powers rather than destroying them. Or maybe, and this is my favourite thought, they do one or the other depending on your ability to critically reflect on what the technology is offering you. ChatGPT, after all, does come with a cognitive health warning that the data it supplies may be wrong. It is up to the wit and discernment of the human using it to sort the right from the wrong. Human intelligence is not quite obsolete, but it does face new demands.
One intriguing hypothesis for the supposed cognitive decline has been advanced by the psychologist Elizabeth Dworak, lead author of the aforementioned 2023 study seeking to explain the negative Flynn Effect. As she told The Hill newspaper: ‘The line can’t go up forever. It’s called the ceiling effect. You eventually hit that threshold.’ Again, this is intuitively plausible. What goes up must come down. Athletes can only run so quickly. Presumably, then, there will come a day when some sprinter holds on to the 100-metre world record, in principle, forever. Similarly, there’s only so far mere human intelligence can go.
But the parallel isn’t a good one. Firstly, world records are measured objectively by clocks. Second, human intelligence is too slippery and culturally relative to yield to objective measurement. What we regard as intelligent or stupid changes over time, just like the skills that we regard as important or irrelevant. Some abilities valorised by IQ tests decades ago may be worthless in 2026.
The grand delusion of IQ scores is that they are objective facts rather than relative measures of intelligence. Despite what hypotheses about the negative Flynn Effect suggest, these numbers do not offer incontrovertible evidence that us poor mugs in the developed West are more stupid on average than our grandparents. That judgment may say more about our witlessness than our intelligence. Or, at least, I hope I’m not stupid in believing so.
