As international data gets more readily available, we see ever-more papers that look at the relationship of inequality with something new. Still, I was slightly taken aback to see a paper – in the Proceedings of the Royal Society (B), no less – that related inequality to preferences among women for masculinized faces.
The story goes like this. A team from Aberdeen/Stirling led by Lisa DeBruine hypothesised that women will prefer more masculinity where health is worse – on the grounds that the benefits of healthy offspring are outweighed by the costs of low paternal investment in the child. They collected their masculinity preferences data online (from 5000 straight white women in their early 20s in 30 countries), and pretty much found what they expected. (I found the image above in a pop-sci report of DeBruine’s research).
An international (Aus-Sweden-UK) team led by Robert Brooks was invited to comment on the paper, and they were the people that decided to get inequality involved in all this. They used DeBruine et al’s data – which was made available on the web – and related masculinized faces preferences to the standard inequality measure, the Gini coefficient. They found amazingly strong links between the two – a model including only the Gini index and an intercept explained 69% of the variance in masculinity preferences.
Brooks et al’s argument is that masculinized faces may ‘signal dominance, which, in certain environments, predicts competitive success in male hierarchies’. In other words it’s not just a signal of health, but also a signal of how well people will cope in competition between men, and that becomes increasingly important in high-inequality societies. To back this up, they show that homicide rates are a stronger mediator between inequality and masculinized preferences that health is.
Back came DeBruine to comment on Brooks et al’s comment. They don’t accept Brooks et al’s argument about homicide rates, which disappear when they control for country wealth (GNP). And while they accept the cross-national findings about inequality, they try and replicate this in a sample of US states, and instead find that health – but not inequality – predict masculinized preferences there.
What does all this mean? Well I’m not going to adjudicate on the particular issue of masculinity preferences, which would take me far away from any subject I can profess any expertise in. It does remind me of the difficulty of doing cross-national research on inequality (cf. Wilkinson & Pickett), which I wrote about back in the very first post on the blog.
But primarily it makes me think that there are a huge number of possible implications of more unequal societies, if Richard Wilkinson’s argument about the psychosocial effects and constitution of society are correct. What I’d really like to see is a systematic attempt to make as many predictions as possible, and then to rigorously test a large number of these to see how well they stand up. My friend – and occasional Inequalities writer – Rob de Vries has done some interesting work on this, which he’ll hopefully put up on the blog later in the year.
In the meantime, I’m filing this under ‘baffling inconclusive findings about inequality’, a box which is getting surprisingly large…