Peterson's 'gender equality paradox' returns to the timeline — and the data behind it is more contested than the clip suggests

A 24 June 2026 post on X, drawn from a Jordan Peterson video, has revived a familiar claim: that the more equal a society becomes, the larger the psychological differences between men and women appear. "Men tend to be more interested in things, women in people," the clip runs. "That's the largest" difference. Within hours the post was being shared in timelines already saturated by a long-running fight over gender, science and what gets called "natural" online.

The claim sounds tidy. It is also older, and more contested, than the clip suggests. The data behind the "things versus people" formulation has been on record since at least the early 2000s, and the most-cited finding in this corner of the literature — that countries with higher gender equality sometimes show larger sex differences in certain personality traits and occupational preferences — has drawn sustained methodological pushback. The version doing the rounds on Wednesday flattens that debate into a slogan. The actual record is messier, and more interesting, than the slogan permits.

What the original finding actually said

The phrase "gender equality paradox" entered wide circulation through a 2018 paper by Gijsbert Stoet, a psychologist then at the University of Essex, and David C. Geary at the University of Missouri. They used data from the World Values Survey and OECD measures of gender equality to ask a specific question: in countries where women have more legal rights, educational access and economic opportunity, do sex differences in certain psychological traits narrow — as a "socialisation-only" account would predict — or persist, or even widen?

Stoet and Geary's answer, for the particular traits they measured, was that differences persisted or widened. The trait that gets quoted most often is the "things-versus-people" axis — an interest pattern in which men, on average across large samples, gravitate toward mechanical or technical objects, and women toward social and caregiving contexts. The effect was largest, by their reading, in the most gender-equal countries.

That last clause is the part that travels. It also happens to be the part most prone to oversell in social-media shorthand.

The replication record, and where it bites

Three lines of subsequent work complicate the clean version of the story.

First, replication. A 2024 preprint and subsequent peer-reviewed work by a separate research group attempted to re-run the Stoet–Geary analysis using both the original World Values Survey items and an alternative dataset. They recovered the headline effect — greater sex differences in more equal countries — for some items, but not for others. The pattern was sensitive to which specific questionnaire items were included, and to how the country-level equality index was constructed. In other words, the "paradox" is real for some operationalisations of the variables, and not for others.

Second, causation. Even researchers broadly sympathetic to the finding have noted that cross-country correlations cannot, by themselves, tell us whether equality produces divergence, whether a third variable produces both, or whether measurement artefacts in equality indices distort the picture. A 2020 paper in the journal Psychological Science showed that the size of the country-level effect shrinks substantially once certain statistical assumptions are changed — without disappearing entirely.

Third, scope. The Stoet–Geary finding concerns psychological traits and self-reported interests. It does not directly support claims about innate cognitive ability, the existence of which is itself a separate empirical question with its own contested literature. Conflating the two is a frequent move in online discourse and a near-universal move in the comment threads under viral clips.

Where Peterson's framing fits — and where it doesn't

Peterson's broader commentary over the last decade has leaned hard on the "things-versus-people" pattern as evidence for what he has called deep, biologically rooted sex differences in vocation and temperament. The clip circulating on Wednesday is a compact restatement of that posture: equal societies do not converge, therefore biology is doing the work.

The argument has internal logic. But the data it leans on is narrower than the rhetoric suggests. The Stoet–Geary effect, even at its strongest, is a population-level correlation measured on specific self-report instruments. It is consistent with several causal stories, only some of which centre innate sex differences. It is also consistent with stories in which equal-opportunity societies simply allow previously suppressed variation to express itself — variation that may itself have social origins further upstream, or that may be partly biological, or both.

The strongest version of the social-constructionist counter is not that differences don't exist; on this point the data is reasonably clear that average differences in interests do exist. The strongest counter is that "the data shows biology is destiny" is a step the data does not, in fact, take. The interesting question is what mix of mechanisms produces the pattern — and that question is, as of mid-2026, genuinely unresolved.

Counter-narrative: what the global picture actually looks like

Outside the Anglo-American culture-war axis, the framing tends to be less charged. Researchers in Scandinavia, where the equality indices sit highest, have produced work that complicates the paradox in a specific way: in those countries, women now outnumber men in higher education across most fields, including several STEM disciplines, while remaining underrepresented in the most math-intensive subfields and in engineering. The pattern is therefore not convergence across the board, and not divergence across the board. It is field-by-field, with some sectors narrowing and others holding.

That finding sits awkwardly with both poles of the online argument. It does not support "biology is everything." It also does not support "equality will erase the gap." It supports something more boring and more accurate: that the gap is contingent, partly on the field, partly on the measure, partly on cultural context, and that no single viral clip can capture the texture of the actual record.

Stakes: why a thirty-second clip matters

The reason the clip travels is not that it contains new information. It is that it offers a tidy answer to a question that has real-world consequences — who ends up in which jobs, how parents read their children's interests, how universities set admissions targets, how employers write job adverts. Each of those domains treats the "men like things, women like people" generalisation as if it were either settled or politically inconvenient, and acts accordingly.

The honest version of the record — that population-level differences in interests exist, that some of them persist or widen in more equal societies, that the cause is genuinely contested, and that policy implications are not derivable from a single correlation — is harder to fit in a thirty-second video. It is, however, the version that holds up under scrutiny. The clip will keep circulating. The data behind it is unlikely to settle in the form the clip presents.

Desk note: Monexus treats this story as a culture-and-research beat rather than a politics beat — the underlying question is empirical, not partisan, and the wire has covered it unevenly. Where Peterson-adjacent material travels on X, this publication will report the finding in its strongest form and the methodological pushback in equal weight.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://en.wikipedia.org/wiki/Gender_equality_paradox
• https://en.wikipedia.org/wiki/Gijsbert_Stoet
• https://en.wikipedia.org/wiki/World_Values_Survey