Originally published 2021-05-26.
A study led by Khandis Blake at the University of New South Wales (UNSW) in Sydney found that sexualised selfies were posted more often in places with high income inequality, and were not linked to gender inequality. The researchers suggested that when the gap between rich and poor is wide, status anxiety rises and appearance becomes one more way to compete. The finding is about places, not individuals, and it rests on keyword searches rather than a look at every image.
The paper, "Income inequality not gender inequality positively covaries with female sexualization on social media", was published online in the Proceedings of the National Academy of Sciences (PNAS) on 21 August 2018, nearly three years before the earliest version of this page appeared.
What the researchers measured
Over one month in 2016, from 8 June to 7 July, the team tracked public posts on Twitter and Instagram worldwide. They collected 453,335 posts that combined one of 31 "sexy" or "hot" terms with one of six "selfie" terms, and filtered out posts with obvious pornographic keywords.
About 68,500 of those posts could be linked to a location, covering 5,567 US cities, 1,622 US counties and 113 countries. The researchers then compared how common these posts were in each place with measures of income inequality, gender inequality and economic development.
To check what the keywords were catching, they inspected a sample of 1,500 posts. Most came from women, and most of those were original selfies of the women themselves. Much of the content posted by men had been reshared from women's accounts.
The main findings
The clearest pattern was with income inequality, measured by the Gini coefficient. In US cities, counties and across countries, places where incomes were more unequal had more sexualised selfies. In US cities, each standard-deviation rise in inequality was linked to roughly a 31 to 34 percent increase in the expected number of such posts, after other factors were taken into account. The cross-national link was stronger among more developed countries.
The researchers also found supporting evidence offline: in US cities, spending at beauty salons and women's clothing stores rose with income inequality too.
Gender inequality told a different story. The associations were small, inconsistent and mostly not statistically significant. The authors summarised it in one blunt line:
"We found no association with gender oppression."
Why this challenged a common assumption
A widely held view in psychology and popular culture is that women's focus on appearance reflects patriarchal pressure: in societies where women are valued mainly for their looks and hold less power, sexualisation is a sign of disempowerment. If that were the main driver, sexualised selfies should be most common where gender inequality is greatest.
The data did not show that. Instead, the posts tracked economic competition. Blake's explanation was that inequality raises anxiety about where people sit on the social ladder, and that appearance can bring real social and economic rewards. In her words:
"Rightly or wrongly, in today's environment, looking sexy can generate large returns."
From an evolutionary perspective, the team argued, competing for status and partners in this way is a rational response to local incentives, not evidence that the women involved are shallow or helpless. Blake urged readers to see a woman posting such a photo as "a strategic player in a complex social and evolutionary game", rather than as vacuous or a victim.
Limitations worth knowing
The paper is careful about its own limits, and they matter when interpreting headlines:
- Places, not people. The analysis is ecological: it compares cities and countries. It cannot say why any particular person posted any particular photo.
- Correlation, not causation. Inequality and sexualised selfies rise together, but the study cannot prove one causes the other.
- Keyword-based. Posts were found by words such as "sexy selfie", not by viewing every image, and only English terms were tracked.
- A narrow window. The data covered one month and two platforms, and only about 104,000 posts had usable location information.
- Mixed robustness. Some cross-national tests were sensitive to how the model was specified, and in some of them the income-inequality effect was not significant.
- Supply and demand. Because men reshared some content, the study cannot fully separate who produces the images from who promotes them.
None of this cancels the finding, but it means the evolutionary explanation, that women are competing for status, is an interpretation of the pattern rather than something the data directly prove.
How the study was reported
Coverage at the time was mostly enthusiastic and rarely mentioned the limitations above. Some headlines suggested women were posting sexy selfies to "combat" inequality, which goes further than the paper. Others described a link with "social anxiety"; the authors' argument was about status anxiety created by inequality, a social rather than a clinical idea.
A fair reading is narrower: across thousands of places, posts labelled as sexy selfies were more common where incomes were less equal, and that pattern held up better than the link with gender inequality.
What to take from it
The research is a reminder that social media behaviour is shaped by the wider economy, not only by personal vanity or by gender norms. It also invites a less judgemental view of how people present themselves online. A photo that looks frivolous may be one small move in a broader competition for attention, opportunity and status that unequal societies tend to intensify.
At the same time, a regional pattern cannot explain any one person's choice. People post selfies for fun, confidence, work, friendship and many other reasons, and the study was never designed to judge individuals.
Sources: PNAS: Income inequality not gender inequality positively covaries with female sexualization on social media; The New Daily (27 August 2018). Earlier coverage of this story appeared on Business Insider Australia.
COMMENTS