Investigating the “STEM gender-equality paradox” – in fairer societies, fewer women enter science

Screenshot 2018-03-14 08.41.46.png
The percentage of women with STEM degrees is lower in more gender-equal countries, as measured by the WEF Gender Gap Index. Image from Stoet & Geary, 2018.

By Alex Fradera

The representation of women in STEM fields (science, technology, engineering and maths) is increasing, albeit more slowly than many observers would like. But a focus on this issue has begun throwing up head-scratching anomalies, such as Finland, which has one of the larger gender gaps in STEM occupations, despite being one of the more gender equal societies, and boasting a higher science literacy rate in its girls than boys. Now a study in Psychological Science has used an international dataset of almost half a million participants that confirms what they call the “STEM gender-equality paradox”: more gender-equal societies have fewer women taking STEM degrees. And the research goes much further, exploring the causes that are driving these counterintuitive findings.

Gijsbert Stoet at Leeds Beckett University and David Geary at the University of Missouri analysed several large and often publicly available datasets, like the gender inequality measures taken by the World Economic Forum (WEF; based on metrics like women’s earnings, life expectancy and seats in parliament) and UNESCO data on STEM degrees.

The researchers found the percentage of women STEM graduates is higher for countries that have more gender inequality. For instance, countries like Tunisia, Albania and Turkey, which come out the poorest on the WEF gender equality measures, see women making up 35-40 per cent of STEM graduates, whereas in countries with more gender equality, like Switzerland and Norway, the figure is lower at around 20 per cent, similar to Finland.

To better understand the STEM gender-equality paradox, Stoet and Geary accessed results from a 2015 OECD educational survey of the science literacy and attitudes to science of 15 to 16-year-old students from 67 countries. Objectively, neither boys or girls were more scientifically literate – girls were better in 19 countries, boys in 22, with no difference in the others.

These survey results suggest it is not girls’ lack of scientific knowledge or negative attitudes toward science that holds them back. It’s possible, however, that girls might match or outperform boys in science lessons in some countries and still be making a rational choice to avoid STEM routes because they outperform boys even more in other areas (it’s well documented that girls outperform boys at school on many topics, on average).

Using the OECD survey, Stoet and Geary calculated a personal ranking for each student of their relative ability across the three main areas of maths, science and reading. In all but two of the 67 nations, boys more than girls were personally stronger in science (80 per cent of boys were personally strongest in either maths or science; in contrast, half of girls were personally strongest in reading). Stoet and Geary found that boys tending to be personally stronger in science was most apparent in more gender-equal countries, the very countries where boys go on to pursue more STEM careers. This smooths some of the kinks in the STEM gender-equality paradox: in fairer societies, boys seem to be optimising their future by pursuing science-like activities, whereas girls have other options on the table.

However, questions remain. For instance, why should a motivated and academically talented teenage girl forego a science route because her reading slightly outperforms her other literacies? After all, reading and writing skills are also beneficial to scientists, from paper writing to fundraising.

Stoet and Geary tried to address this question by focusing on gender differences in motivation, in terms of interest, confidence and enjoyment. In 60 per cent of countries, boys showed more interest in science than girls (in various fields, from disease prevention to energy), and the gender-gap in scientific interest was greater in the most gender-equal countries; boys also expressed more confidence in their science abilities in 39 of 67 countries, especially the gender-equal ones. And even though girls reported enjoying scientific activities more than boys in two-thirds of the countries, boys’ enjoyment was higher in the more gender-equal countries.

Why are boys most enthusiastic, interested, and personally strongest at science in more gender-equal societies? The authors suggest that in highly stable countries with strong welfare systems, people can pursue their calling and unlock their personal potential, building their future around their genuine interests and personal strengths. This echoes the finding popularised in online lectures by psychologist Jordan Peterson that sex-related personality differences are higher in gender-equal societies – when societies’ social pressures are less tyrannical, individual tendencies can be expressed more freely.  In more repressive cultures, by contrast, young people are liable to prioritise pragmatism – food on the table – over self-actualisation, and as STEM jobs tend to be stable and well-paid, that would encourage more female representation. Consistent with this, Stoet and Geary used a United Nations life satisfaction measure as a proxy for cultural stability and found that more women took STEM degrees in countries where life satisfaction is lower – which tended to be in the unequal societies.

These findings suggest we need a nuanced approach to the sticky issue of gender and participation in science. Firstly, there is no question from this data that objectively, young women across the globe are just as capable to tackle scientific subjects as are young men. And even after taking into account the gender differences in science attitudes and personal strengths, the researchers calculated that, in a society where women’s rational preferences led directly to their level of STEM participation, we should see women take 34 per cent of STEM degrees, while the actual global average is 28 per cent – so other factors unaddressed in this study are clearly leading women away from science roles. So the study doesn’t suggest that we sit on our laurels and validate the status quo.

It does suggest, however, that we misunderstand the current level of STEM gender imbalance if we attribute it entirely to social injustice. A substantial cause of the current STEM gender mix may be the product of young men and women making considered, rational choices to leverage their strengths and passions in different ways.

The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education

Alex Fradera (@alexfradera) is Staff Writer at BPS Research Digest

20 thoughts on “Investigating the “STEM gender-equality paradox” – in fairer societies, fewer women enter science”

  1. I welcome this study based as it is on real- life , non- experimental data. Human animals are biopsychosocial beings and no behaviour can be analysed in isolation from any one of these spheres of influence with external validity. After all, everyone alive today are the results of the greatest experiment of them all (I am NOT advocating the existence of a ‘creator’ when making this comment).

    I think after having explored the concept of ‘control’ and self- determinism, that the results might well be explained in terms of ” when people feel they have choice, gender- role differences cease to be so important for a healthy self – identity”. We are so concerned with ‘political correctness’ and issues of ‘equality’ for everyone,that we forget that we are first and foremost, animals just like any other, products of thousands of years of evolutionary adaptation to variable geographical, tribal, cultural social orders.

    Psychologists need to understand that they represent a tiny minority of the total population of the world and that not everyone thinks about equality in the same positive way we do. Neither are they necessarily as bothered about it as Western, Educated, Industrialised, Religious, Democratic (WEIRD,Norenzayan) cultures.

    I propose that even within Western cultures, not every woman wants to enter the sciences or values her career over being a supportive wife and mother and that I am fed up with nurturing, cleaning and caring- roles being consistently undervalued no matter what the gender or ethnicity of the person doing them.

    Personally, I think this is to do with the universal emotion of disgust and the association of these roles with dealing with bodily fluids/ faeces, death and ‘contaminants’. I think it is the reason why people can’t be bothered to clean up their dogs faeces and make every effort to discard it as soon as possible, preferably anonymously, without carrying it to the nearest disposal bin.

    I also firmly believe that this emotion, not gender or ethnicity, that defines the nature of inequality in all society’s e.g. the caste system, sewage pipeline and toilet cleaners, nurses, refuse disposal, anyone in fact who has to clean up ‘mess’ – mental and physical.

  2. What is desperately lacking in the majority of research on anything is cultural awareness.

    The phenomenon described here has been around for a long time for anyone who cares to look. In European countries men are still regarded as superior to – including smarter than – women, tho less so than in some other cultures. What is present in European cultures but absent many other places is the belief that maths is the true pure measure of intelligence. We are also prone to inbuilt talent models of success at some fields of endeavour, particularly STEM. This leads to the belief that men are ‘naturally’ better at maths and maths based disciplines. And as women are less likely to attempt study in areas regarded as requiring natural talent for success.

    That STEM disciplines are not regarded as a measure of natural ability in many cultures leads to these being okay for women to study successfully.

  3. What is desperately lacking in the majority of research on anything is cultural awareness.

    The phenomenon described here has been around for a long time for anyone who cares to look. In European countries men are still regarded as superior to – including smarter than – women, tho less so than in some other cultures. What is present in European cultures but absent many other places is the belief that maths is the true pure measure of intelligence. We are also prone to inbuilt talent models of success at some fields of endeavour, particularly STEM. This leads to the belief that men are ‘naturally’ better at maths and maths based disciplines. And as women are less likely to attempt study in areas regarded as requiring natural talent for success they avoid these areas.

    That STEM disciplines are not regarded as a measure of natural ability in many cultures leads to these being okay for women to study successfully.

    1. I wonder if changes are on the way in the USA (and perhaps other countries)….ecology courses at the university level have been attracting a greater percentage of women students than in most other science courses – biology also to a lesser extent. Hence, many universities in the USA are now developing more ecology-based courses in part to reduce the gender gap in science.

  4. A brief note on the intellectual history of one of the important ideas in this debate.
    Over the years, much has been made of the fact that women do as well on mathematical tests as do men, as if it were a direct implication that in an unbiased world there would be as many women in mathematical fields as there are men. See, for example, the well known debate between Steven Pinker and Catherine Spelke (
    More recently people have recognised that personal choice depends on a comparison of alternatives. My decision to enter a STEM field does not depend on my abilities in math relative to others (although this will affect, say, whether I become an electrical engineer or an electrician). My decision to enter a STEM field depends on how good I am at math relative to how good I am at other things. It’s a personal choice. What is the best option for me?
    This idea first appeared about 200 years ago in the works of David Ricardo in his book “The Principles of Political Economy and Taxation (1817)”. He was concerned with the reasons why countries (rather than people) specialize in one commodity or another but the principle is the same. If you ask people why Canada exports softwood lumber to the United States, most people might say “Because trees grow really well in Canada.”. But softwood trees grow terribly in Canada – it is far too cold for much of the year. Trees grow much faster in North Carolina. The reason Canada specializes in growing trees is that there is precious little else that can be done with the land.
    In Economics the concept is called “comparative advantage” and the implications run quite deep. It’s not just that it helps to explain occupational choice. It also tells you that there is a cost to encouraging more women into STEM fields. If a woman with excellent technical and social skills has the potential to be an effective doctor and we encourage her to enter Computer Science, our society may be losing more with the loss of a physician that we are gaining with the addition of another software engineer. As well, if we are nonetheless committed to a more even distribution of genders across occupations, we might accomplish this as easily by encouraging boys to read. This might be better than teaching girls to code because it will also help to address the general under-representation of boys in higher education.
    I address these issues in a bit more detail here:
    So far as I know, the first paper to look at the impact of relative grades on the choice of university programme is here:

  5. Apparently, in fairer societies, women make use of the potential to “express their individual tendencies more freely”. However, it seems that in these “fairer” societies, men are still being pressured into a provider role and do _not_ have that freedom, as it is expected from them to end up in a well-paid position, even if it means deciding against these individual tendencies if they do not result in good pay or status.

  6. The article’s conclusions are based on the assumption that the overall gender-equality index (OGEI) equates to the degree of equality in attitudes to women entering STEM employment. Attitudes toward the acceptability of women entering certain areas of employment have not been shown to be related to the OGEI but discussion of the so-called paradox assumes this.

    Closer examination of the factors included in the OGEI show that these are not generally related.

    Furthermore, the chosen indicators are of necessity imprecise. For instance, women’s political empowerment looks only at the highest levels (parliaments, top governmental positions, and years with a female head of state). It then gives a weight to each subfactor to arrive at an overall figure between 0 and 1. The method of calculating weights sounds plausible but arrives at 44.3% for having a female head of state. It is debatable, to say the least, whether the calculation of gender equality in political empowerment should very largely depend on the historical accident of having a female head of government.

    Other subfactors have similarly debatable weightings while the indicators have widely differing ranges and distributions.

    So many factors (historical, cultural, demographic etc.) are left unexamined in this comparison that the OGEI is useless in examining reasons for STEM gender equality.

    1. Hi Les

      The fact that the OGEI does not closely map onto attitudes to STEM doesn’t seem all that problematic for the claims I’m reading here, which are:

      (1) people seeking to self-actualise tend to want to cultivate existing talents “their gift to the world” or “their road to posterity”, however one wants to see it
      (2) in more gender-equal societies, people are more likely to pursue self-actualisation, as fewer pressures exist to call on them to compromise towards more practical goals.
      (3) the tendency for girls and young women to show a strong aptitude in language/reading therefore becomes more pronounced in a gender-equal society.

      As far as I can see, these claims remain regardless of how much the Gender-Equal measure focuses on STEM. I can even see a STEM-focused measure capturing something very different from what (2) is about.

      The more general claim that the OGEI is just not very good, I don’t have enough insight to comment on. Are there better indicators of overal gender equality out there?

      1. Hi Alex,

        Your argument seems to me to amount to the following. Please forgive me if I have misunderstood.

        You seem to assume that, in more gender-equal societies, people are more likely to pursue self-actualisation and, therefore, females will tend to go into occupations where language/reading are important. You seek to show that this explains the gender imbalance in STEM employment as follows. You assert that there is a “tendency for girls and young women to show a strong aptitude in language/reading.” In more “gender-equal” societies, they will therefore go into areas where these skills are useful.

        Your assumptions are (1) that OGEI informs us accurately which societies are more gender-equal; (2) that women show much stronger abilities in verbal (Language/reading) skills; (3) that these skills are of less value in STEM subjects; (4) that in more gender-equal societies, people have the economic freedom to choose their occupations according to their wishes.

        (1) I have explained how the OGEI is a very unreliable instrument. You say that you don’t have the insight to comment on this but the link I provided gives an explanation of how the ratings are arrived at. Not only have you not critically examined their methods, but no one who has blithely cited these findings has either. You are in very good company!

        What would be more useful would be to actually examine the prevailing views and prejudices in the so-called more GE societies. Just to examine the graph at the top of the article would be quite informative. Ireland (only recently emerged from rural poverty and where abortion is illegal) is the 4th most equal society? Switzerland, where women have only had equal voting rights for 40 years, is 5th? Finland (1st) is hardly typical, having gone through rural poverty, wars, occupation, loss of territory, and Soviet pressure until about 40 years ago. There are plenty more anomalies (and I wonder where Cuba would come).

        (2) Such evidence as there is shows a SLIGHT tendency to exceed the male mean in verbal skills (and a slight tendency the other way in visuo-spatial skills). This was dealt with by the APA in an online article in 2014 (

        (3) Now this is an n = 1 observation but, in my extensive education and training, as well as teaching, in the “hard” sciences of biochemistry and physics, I have found verbal skills (reading and language) to be absolutely crucial.

        (4) People need to earn a living but this fact is almost totally ignored by those who hail these findings as evidence that…what…women don’t want to work in STEM?



  7. I have a theory relating to the STEM gap, which I’ll outline here:

    1. The majority of teachers at primary and secondary school are female. As females, they are more likely to show favouritism towards girls.

    2. Although the female teachers may tend to favour girl students, the objective marking schemes used in STEM subjects means that students can still get objective feedback on their ability during tests. Since males and females are of broadly similar ability in STEM subjects, the objective marking results in equal proportion of male and female students thinking they have a talent for STEM.

    3. In contrast, marking in non-STEM subjects tends to be more subjective. Although male and female students have broadly similar ability in non-STEM subjects the subjective marking (from teachers biased toward girls) makes far more female students think they have a talent for non-STEM subjects.Furthermore, in mixed schools the ‘top student’ in each class for non-STEM subjects is almost always female (in terms of the teachers’ opinion).

    4. When choosing A-level and degree course, students will naturally go for those subjects they think they have a talent for. As a result, most girls choose non-STEM subjects because they think they have a talent for them, whereas many boys choose STEM subjects, partly because they think they do not have a talent for any non-STEM subjects.

    5. This is exacerbated by teachers not giving realistic information about job prospects, careers and salaries associated with the different subjects. This means that girls are not given enough incentive to choose STEM subjects (which generally lead to well-paid careers) over non-STEM subjects (which generally do not).

    In conclusion, one root cause of the STEM gap is teachers erroneous belief that girls tend to be better than boys at non-STEM subjects.

  8. Instead of sugar coating it, how about this? In countries with higher levels of social security which are also somewhat more gender-equal, misogyny and patriarchy still exist. Therefore women are more inclined to choose careers that expose them to less harassment and mistreatment. These are careers that are not STEM.

    1. HI Renee

      I wasn’t sugar coating it, I simply reached a different conclusion from you. If yours turns out to be stronger I’d be happy to endorse it.

      This is the claim made by the paper as I understand it:

      Claim: There are gender-based occupational preferences that would on average influence women’s job decisions if unconstrained.
      If so, we would see more gendered divergence in job choices in less constrained environments
      We do see more gendered divergence in job choices in less constrained environments
      Therefore, the claim has gathered some evidence.

      I’d like to phrase your argument in my own words to have a better chance of understanding it.

      Claim 1: Women will make choices to avoid harassment and mistreatment if unconstrained (by eg the need to feed their family)
      Claim 2: Taking STEM positions exposes you to more harassment and mistreatment versus other jobs
      If these are true, we would see women avoiding STEM positions in less constrained environments
      We do see women avoiding STEM positions in less constrained environments
      Therefore, the position has gathered some evidence.

      Is this a fair summation?

  9. I read your article and it was very interesting about International Conferences but i’ve seen a best conferences website. They work for their passion and International Research Conference, Publish your research, Research paper publication, Conference and publications, Conference on management and social sciences, Conference on Technology, Academic research conference, Multidisciplinary Studies, Must visit for a once. Best in the world

Comments are closed.