Gen Z IQ: Is Gen Z Less Intelligent Than Its Parents? What the Data Show

We found no study that compares the IQ of Gen Z, people born from 1997 onward in Pew Research Center's definition, with its parents' on the same test. The claim that Gen Z has a lower IQ rests on one spoken sentence at a US Senate hearing. On 15 January 2026 the neuroscientist Jared Cooney Horvath said Gen Z is the first generation in modern history to underperform its parents on basically every cognitive measure, including "even general IQ", and gave no figure for it. The evidence behind the claim is narrower: reading and maths scores of teenagers have fallen in many countries since the early 2010s, and conscript IQ in Norway peaked for men born in 1975 and fell afterwards. But the best IQ data stop with men born in 1991. This guide sets the claim beside the primary sources.

What did the Senate witness actually say?

The hearing was "Plugged Out: Examining the Impact of Technology on America's Youth", held by the Senate Commerce Committee on 15 January 2026 and printed as S. Hrg. 119-472. Horvath's spoken statement began: "Our kids are less cognitively capable than we were at their age." He said every generation since the late 1800s had outperformed its parents "until Gen Z", and that Gen Z "is the first generation in modern history to underperform us on basically every cognitive measure we have, from basic attention to memory to literacy to numeracy to executive functioning to even general IQ". He asked what happened "around 2010" to separate schooling from cognitive development, pointed to classroom technology, and added: "of course, this is all correlative." He named no dataset for the generational claim.

His eight-page written testimony is more cautious than the speech. We searched its text and found no "Gen Z", no "first generation" and no "standardized"; "IQ" and "Flynn" appear only inside the titles of references. It dates the turning point to the mid-2000s, saying the long rise in cognitive performance "plateaued then reversed in many Western nations". The psychologist Jean Twenge, speaking on the same panel, dates the fall in US 8th and 12th grade reading and maths scores to after 2012. So the decline starts in the mid-2000s, around 2010 or after 2012, depending on which sentence of the hearing you read.

Press coverage widened the gap. Fortune reported that Horvath told the Senate, in written testimony, that Gen Z is the first generation to score lower on standardized tests than the one before it. That sentence is not in the written testimony, and the phrase "standardized tests" appears in neither of his statements. Fortune's remark that literacy and numeracy tests "aren't always indicative of intelligence" is also the journalist's caveat, not the witness's. Weeks later, in a newsletter post of 6 February 2026, Horvath wrote that "we cannot concretely say that Gen Z is less intelligent than Millennials", and described IQ as a measure of "school-ability". The sensational version of the claim is therefore stronger than the one its author defends.

What do school-test scores show for Gen Z?

Here the numbers are real, but they come from achievement tests. PISA tests 15-year-olds, so the 2022 cycle covered students born in 2006 and 2007 and the 2025 cycle those born in 2009 and 2010 (our arithmetic; Pew puts the Millennial and Gen Z boundary at 1996 and 1997).

• PISA 2022 OECD averages: 472 in mathematics, 476 in reading and 485 in science
• PISA 2025 OECD averages: 463 in mathematics, 461 in reading and 482 in science; mathematics is down 22 points and reading 28 since 2015 (see our guide to PISA 2025)
• NAEP long-term trend, age 13, 2023: mathematics down 9 points and reading down 4 from the 2019-20 assessment, and down 14 and 7 points from a decade earlier

PISA is scaled to a mean of 500 and a standard deviation of 100, so a 22-point fall is 0.22 of a standard deviation, about 3.3 points on an IQ-style scale with a standard deviation of 15. That conversion treats an achievement test as if it were an IQ test. Our guide to the PISA test versus an IQ test explains why the two are not interchangeable, and our guide to screen time and test scores in children covers the screen-time evidence.

What do IQ-type tests show?

The best evidence is Norwegian conscription testing. Bratsberg and Rogeberg (PNAS, 2018) analysed 736,808 valid scores from men born between 1962 and 1991, converted to an IQ scale with a mean near 100. The average rose from 99.5 for the 1962 cohort to 102.3 for 1975, then fell to 99.4 for 1989 and 99.7 for 1991. Their selection-corrected model puts the 1975 to 1991 decline at about 0.33 points a year within families (interval 0.26 to 0.40), roughly 3.3 points a decade. The same rise and fall appears between brothers, which points to something in the environment, not to who was having children. The authors say they cannot identify the causal structure.

Denmark is similar but smaller. Teasdale and Owen (Intelligence, 2008) found that Danish conscripts gained modestly on a four-test battery between 1988 and 1998, then lost about 1.5 IQ points by 2003 to 2004, among men pursuing higher education and those who were not. In the United States, Dworak, Revelle and Condon (Intelligence, 2023) followed 394,378 online adults from 2006 to 2018: composite, matrix-reasoning and number-series scores fell, verbal reasoning barely moved, and three-dimensional rotation rose. Online volunteers are not a random sample, and these are adults. A 2016 review by Dutton, van der Linden and Lynn found nine studies reporting negative Flynn effects in seven countries.

Set against that, the century-long picture is still one of gains. A meta-analysis of 271 samples and almost 4 million people (Pietschnig and Voracek, 2015) found full-scale gains of 0.28 IQ points a year between 1909 and 2013, stronger for adults than for children and smaller in recent decades. Even the newest US Wechsler comparison shows a smaller gain, not a loss: in two counterbalanced studies (186 people aged 16 to 90 on the WAIS-IV and WAIS-5, and 98 sixteen-year-olds on the WISC-V and WAIS-5), the Flynn effect came out at 1.2 points where about 3 had been expected (Winter, Trudel and Kaufman, 2024). The samples were small and practice effects were possible. Our guide to the Flynn effect stalling sets out why norms age.

Why are "first generation" and "IQ" the weak links?

Three gaps stand out. First, no dataset we read ends in Gen Z: the Norwegian series stops with men born in 1991, the Danish men were tested between 1988 and 2004, and the US online sample is adults. Second, the school-test falls are real but measure reading, maths and science, which is why Horvath's own list runs from "attention" through "literacy" to "IQ" without separating them. Our guide to achievement tests and IQ tests explains the difference. Third, "first generation" is a comparison across birth cohorts in separate countries and tests, not a measured contrast between Gen Z and its parents.

The "six to eight points per decade" figure traces to Horvath's newsletter, which says that "across a handful of well-designed studies" the drop has been "as much as" that, but names none of them. The Norwegian estimate, the one we could verify, is about half of it. We could not identify the 2025 paper called "Digging In" that the newsletter cites for the finding that visually guided tasks held up while memory-heavy ones fell.

What does newer evidence say about attention?

The written testimony cites a 2024 meta-analysis of the d2 attention test (Andrzejewski, Zeilinger and Pietschnig, 287 samples, 21,291 people, 32 countries, 1990 to 2021) as support for stagnation or decline. Its title reports better test performance, and its abstract finds moderate gains in adults but not in children, with more errors alongside faster processing in children. A paper published on 7 October 2026 (Lesigang, Vetter, Voracek and Pietschnig, in Intelligence and Cognitive Abilities) compared Austrian adults tested between 2000 and 2004 with matched adults tested in 2025. Selective attention improved slightly (d = 0.12), interference control was unchanged, and sustained attention may have declined (d = -0.27). The authors say the results contradict the notion of large changes in attention-related abilities in the digital age. The participants were adults aged 16 to 91, not Gen Z.

Causes remain open. Horvath argues that classroom screens are responsible; the Norwegian authors find an environmental cause but cannot say which; other proposals include changes in schooling, media exposure, nutrition and health. Our guide to cognitive offloading and our blog guide on whether AI makes you dumber cover one newer strand of the argument.

What does this mean for your own IQ score?

Nothing changes on an individual report, because every IQ test is re-normed so that the average is 100, as our guide to whether IQ changes over time notes. A drift of 3 points a decade, if it applies to you at all, matters mainly for old norms. Read a score as a rank against people tested on the same norms, and use the percentile calculator to see what the rank means.

To compare any score with the population, take the IQ Metrics test and read the result against the IQ bell curve.

Common questions