If you’ve listened to any Jordan Peterson, you may have heard him bring up the Gender Equality Paradox: countries with higher levels of gender equality (the Nordic countries being the most often used example) often have fewer women in STEM. This seems counterintuitive - many people would expect that countries that had more gender equality would create more opportunities for women to go into fields that they have been historically underrepresented in. There is actually more to the paradox than this - some studies have shown that there are bigger differences in the big five personality traits in more gender equal countries, that the sexes have more divergent preferences in more gender equal countries, and so on.
The main claim that Peterson makes comes from a 2018 paper by Stoet and Geary. Here’s a quick summary of the paper:
Girls outperformed boys in science in 28.4% of countries, whereas boys outperformed girls in 32.8% of countries - the remaining countries did not show a significant difference in academic achievement between boys and girls.
In all countries except for Lebanon, boys’ intraindividual strength in science was larger than that of girls, and in all countries girls intraindividual strength in reading was larger than that of boys. Essentially, even if boys are not better at science than girls in absolute terms, they have a sort of comparative advantage because boys are more likely to be better at science than they are at reading, whereas girls are more likely to be better at reading than they are at science.
The sex difference in intraindividual strength was larger in countries with more gender equality. Take Finland: even though it was the case that girls did better than boys in science, they performed even better in reading, and so girls were less likely to do STEM than in other countries even though they had an absolute advantage in science. The above plot shows the negative correlation between the % of women in STEM, and the GGGI metric of gender equality.
Boys were also more likely than girls to say that they enjoyed science in more gender equal countries, and boys had higher levels of science self-efficacy (how good they thought they were at science) relative to girls in more gender equal countries.
So far, the story fits - the paper argues that the mechanism behind this apparent paradox may lie in the fact that more gender equal countries also tend to have more robust social safety nets, meaning that people need to worry less about paying the bills and ensuring they have a cushion, and can just do whatever they’re interested in or are better at. Essentially, the smaller financial costs of forgoing a STEM route mean that boys and girls are more likely to pursue their intraindividual strengths.
Time for objections!
The above is the story told by Jordan Peterson - but there are a few questions raised about the study by various academics that I think should make us slightly wary of drawing such a strong conclusion just from Stoet and Geary. The most powerful objection is made in Richardson et al (2020) - they claim that the negative correlation between gender equality and the number of women graduating with STEM degrees is contingent on the measurement of equality that is used. Specifically, Stoet and Geary use the GGGI, but Richardson et al argue that GGGI is a worse measure than a different measure (BIGI). The reasons they give are fairly unconvincing, but using another measure is a good way to check how robust the correlation really is. Interestingly, BIGI was actually created by Stoet and Geary because they thought that GGGI was flawed. They write:
The Global Gender Gap Index is one of the best-known measures of national gender inequality, used by both academics and policy makers. We argue that that this measure has a number of problems and introduce a simpler measure of national levels of gender inequality.
As can be seen in the above plot from Richardson et al, the negative correlation between the gender equality metric and the number of women in STEM disappears when using BIGI rather than GGGI (p = 0.518).
And then another objection is made by Breda et al (2020). The argument is this: weirdly, stereotypes about men being better than women at maths seem to be stronger in countries with more gender equality. The argument that Breda et al make is that the negative correlation between gender equality metrics and the number of women in STEM goes away when you control for the stereotypes about men and women in STEM.
Stoet and Geary strike back!
In 2020, Stoet and Geary wrote a reply to the Richardson critique: ‘The Gender-Equality Paradox Is Part of a Bigger Phenomenon: Reply to Richardson and Colleagues’, in which they basically make two claims: the first one is that GGGI is a better metric than BIGI when doing this particular analysis. They write that BIGI is just a simplified metric that shouldn’t be used in this instance because it doesn’t take into account the political empowerment of women, and political empowerment is likely to play an important role in the gender equality paradox. I don’t find that particularly convincing, but make of it what you will. The second point raised by Stoet and Geary is that the finding with regards to STEM is merely one part of the Gender Equality Paradox, and that there are other studies that show that personality differences between men and women are larger in more gender equal countries and that there are larger differences in personal interests between men and women in more gender equal countries.
What to make of all this? The literature on the Gender equality paradox is still developing, and it isn’t clear yet which side is ‘winning’. I don’t find Stoet and Geary’s response to Richardson et al particularly convincing, but similarly I don’t find Richardson et al’s view that BIGI is superior to GGGI particularly convincing. If I had to offer a few take-aways they would be these:
Whether or not the number women in STEM negatively correlates with levels of gender equality is contingent on which measure of gender equality you use, and calculation you use about how many women there are in STEM.
Two contradictory criticisms of Stoet and Geary seem to have emerged: one argues that there is simply no gender equality paradox when you use the right metric of gender equality, and the other is that there is a gender equality paradox, but the paradox is explained by countries with more gender equality being more likely to stereotype women as worse at maths (rather than being a difference in intraindividual strengths).
The literature on the Gender Equality Paradox is still pretty new, and anything we say about it should reflect that, and a rational person is likely to be uncertain about the correlation between gender equality and the number of women in STEM, rather than firmly taking one view or the other.
Interesting, the BIGI seems to be a simplified version of the GGGI, which doesn't look at as many measures. The BIGI only considers, 1) childhood educational opportunities, 2) healthy life expectancy, and 3) life satisfaction. The GGGI considers 14 different measures, weighs them, and then sums them. I figure the good way to help clarify this disagreement would be to separately measure the correlation between the gender-STEM value and each of the 14 measures.
Great piece - alas there are few 'rational' people on the internet!