The Blog

2021-

2021-04-02 e
WRONGTHINK V

Men, women and STEM: Why the differences and what should be done?

Abstract

It is a well-known and widely lamented fact that men outnumber women in a number of fields in STEM (science, technology, engineering and maths). The most commonly discussed explanations for the gender gaps are discrimination and socialization, and the most common policy prescriptions target those ostensible causes. However, a great deal of evidence in the behavioural sciences suggests that discrimination and socialization are only part of the story. The purpose of this paper is to highlight other aspects of the story: aspects that are commonly overlooked or downplayed. More precisely, the paper has two main aims. The first is to examine the evidence that factors other than workplace discrimination contribute to the gender gaps in STEM. These include relatively large average sex differences in career and lifestyle preferences, and relatively small average differences in cognitive aptitudes – some favouring males, others favouring females – which are associated with progressively larger differences the further above the average one looks. The second aim is to examine the evidence suggesting that these sex differences are not purely a product of social factors but also have a substantial biological (i.e. inherited) component. A more complete picture of the causes of the unequal sex ratios in STEM may productively inform policy discussions.

Keywords
discrimination, equality, gender, sex differences, STEM

Never has the issue of gender disparities been as widely discussed, or as bitterly contested, as it has been in recent years. From the Oscars to the political podium, from TV shows to the workplace, disparities are identified and debate inevitably ensues. In the occupational realm, one of the primary focuses of this debate has been the differential representation of men and women in STEM (science, technology, engineering and maths; see Box 1). This was epitomized by the infamous ‘Google memo,’ in which then-Google employee James Damore (2017) questioned the extent to which observed gender disparities in STEM are a product of workplace discrimination. The memo, and Damore’s subsequent dismissal from Google, provoked a great deal of discussion and debate about the causes of STEM disparities and the origins of human sex differences. Unfortunately, much of this debate was decidedly inaccurate in its presentation of the research on the topic. A great deal was said about bias and discrimination, but relatively little about other factors contributing to STEM gender gaps (e.g. Chachra, 2017). Furthermore, to the extent that other factors were mentioned – factors such as average sex differences in academic interests – these were typically attributed to socialization, rather than to biology or to a complex interaction between biological and sociological causes (e.g. Campbell, 2017; a notable exception is Eagly, 2017).

The goal of this paper is to redress the balance. We do not aim to provide a complete survey of the literature on sex differences in STEM; to do so would require a book-length treatment of the topic. Our goals are much more modest. The first is to argue that gender gaps in STEM are shaped to an important extent by factors other than workplace discrimination, including sex differences in preferences, aptitudes and within-sex variability. The second is to argue that these sex differences are not due solely or primarily to learning, socialization or culture. Biology matters as well.

Critics might respond that no one claims otherwise, and that to suggest that they do is merely to attack a straw person. We defend our emphasis, however, on three main grounds. First, it is far from clear that the only people rejecting a significant role for biology in shaping STEM gender gaps are made out of straw. As the psychologist Alice Eagly (2018) has noted, many feminist psychologists have rejected a role for biology in shaping any psychological sex differences. Second, although few experts explicitly deny that biological factors contribute to STEM gender disparities, these factors are often downplayed or ignored. Wang and Degol (2017), for instance, suggest that, although biological factors cannot be ‘definitively dismissed,’ socio-cultural factors are a more likely explanation (p. 123), and Cheryan et al. (2017) do not even mention biological factors in their analysis of the causes of the gender gaps in STEM. It would be easy for non-experts and policy makers to get the impression that, according to many experts, biology is essentially irrelevant. Third, even if everyone did agree that biological factors make a significant contribution (over and above simply encoding the effects of experience), it would presumably still be appropriate to make the case for this position, rather than simply accepting it in the absence of arguments and evidence.

We divide the paper into six main parts. First, we survey the research suggesting that men and women differ, on average, in their career and lifestyle preferences, and argue that these differences are due in part to biological influences. Second, we consider the possibility that men and women differ, again on average, in certain cognitive aptitudes – that men, for instance, score somewhat higher on most tests of spatial ability, whereas women score somewhat higher on verbal tests. Third, we look at the controversial suggestion that men are more variable than women in cognitive ability, such that there are more men at the top of the ability distribution, and more men as well at the bottom. Fourth, we look at the issue of gender discrimination, and argue that, although discrimination plays a role in shaping STEM gender gaps, it plays a smaller one than people often assume, and sometimes favours women rather than men. Fifth, we look at how the arguments and evidence in the first four sections might inform the discussion of policy interventions aimed at addressing STEM gender gaps. Sixth and finally, we consider whether the ultimate aim of such interventions should be to eliminate sex differences in STEM, or simply to eliminate bias and barriers, then let the cards fall where they may.

Sex differences in preferences and priorities

To begin with, we examine arguably the most important contributor to the differential representation of men and women in STEM: sex differences in career-relevant preferences. Specifically, we look at sex differences in interests and occupational preferences and sex differences in life priorities. Having sketched an origins-agnostic outline of these differences, we then make the case that biological factors play an important part in shaping them, and speculate about the evolutionary pressures that might have helped shape the biological contribution.

Interests and occupational preferences

A large literature in psychology shows that men and women differ, on average, in the kinds of occupations that interest them (Konrad et al., 2000; Morris, 2016). One of the most important recent papers on this topic was a comprehensive meta-analysis by Su et al. (2009). The paper focused on two main areas: occupation-relevant interests (e.g. interest in people vs. things) and preferences for specific STEM careers (e.g. engineering vs. mathematics). In both cases, the authors found substantial sex differences that, regardless of their causes, plausibly go some way towards explaining observed STEM gender gaps.

Occupation-relevant interests. Starting with occupation-relevant interests, by far the largest sex difference was that for interest in things (i.e. objects, machines or abstract rules) vs. interest in people. Members of both sexes can be found at every point on the things vs. people continuum; however, more men than women exhibit a stronger interest in things, whereas more women than men exhibit a stronger interest in people. Averaging across studies, Su et al. (2009) found an effect size of d¼0.93 for the people vs. things sex difference. This is notably larger than most human sex differences (Hyde, 2005; Lippa, 2010; Stewart-Williams & Thomas, 2013a, 2013b), and indeed than most effects in psychology (Eagly, 1995). To get an intuitive sense of the magnitude of the difference, if one were to pick pairs of people at random, one man and one woman, the man would be more things-oriented than the woman around 75% of the time.

The people vs. things sex difference immediately suggests an explanation – or rather a partial explanation – for the fact that men outnumber women in fields such as physics, engineering and mathematics, whereas women are at parity with or even outnumber men in psychology, the social sciences and the health sciences: the former fields are of interest to more men than women, and the latter to more women than men, and people tend to gravitate to fields that interest them most (Diekman et al., 2017; Yang & Barth, 2015).

Preferences for specific occupations. Research looking at preferences for specific occupations leads to a similar conclusion. As Su et al. (2009) report, males on average express considerably more interest than females in engineering (d¼1.11), and somewhat more interest in science and mathematics (d¼0.36 and 0.34, respectively). These differences are present by early adolescence and closely match the observed numbers of men and women working in the relevant fields. Su et al. (2009) point out that, if we make the reasonable ballpark assumption that people working in a given field tend to come from the 25% of people most interested in that field, sex differences in occupational interests would account for the entirety of the engineering gender gap and much of the gap in science and mathematics. In short, sex differences in occupational and academic preferences are far from trivial, and plausibly make a substantial contribution to observed occupational gender gaps.

Life priorities

Gender gaps in STEM – and especially in the higher echelons of STEM – may also be shaped in part by average sex differences in life priorities. As with occupational preferences, people vary a lot in their life priorities, and the full range of priorities can be found within each sex. Nonetheless, some priorities are more common among men than women, and others among women than men (Bolotnyy & Emanuel, 2019; Hakim, 2005, 2006; Konrad et al., 2000; Schwartz & Rubel, 2005). One longitudinal study found, for instance, that among adults identified as intellectually gifted in early adolescence, the average man reported placing more importance on career success and income than did the average woman, whereas the average woman reported placing more importance on work–life balance and making time for one’s family and friends (Benbow et al., 2000; Lubinski et al., 2014). These differences were particularly pronounced among people with children, apparently because women’s priorities shifted after they became mothers (Ferriman et al., 2009). Moreover, sex differences in self-reported priorities were evident in real-world behaviour. As Lubinski et al. (2014) observed, for instance, over the course of the last 15 years, the men in their sample spent an average of 51 hours a week doing paid work, whereas the women spent an average of 40.

Of course, sex differences in lifestyle preferences do not explain why the sex ratio is so much more male-biased in maths-intensive STEM fields than in most others. Still, the differences do plausibly help to explain the fact that, in STEM and elsewhere, men outnumber women among the minority in the higher echelons: rising to the top is a priority for fewer women than men, and thus fewer women than men are willing to make the sacrifices required to achieve that goal. To be clear, some women are willing to make those sacrifices, and the majority of men are not. However, more men than women are willing, and this is plausibly part of the reason that the sex ratio at the top is so often male-biased. Note that, according to one large US study (N 4000), the sex difference in career-mindedness is not a result of women thinking that career advancement is impossible for them. The average woman views advancement as just as achievable as the average man, but as less desirable (Gino et al., 2015).
[ ... ]
A mixed picture

In summary, it seems fair to say that the evidence for gender discrimination in STEM is mixed, with some studies finding pro-male bias, some finding the reverse and some finding none at all. What should we conclude? In our view, there are two main interpretations. The first is that the apparently mixed findings are not in fact inconsistent. Rather than there being uniform bias against women, or uniform bias against men, there are pockets of bias against both sexes (and presumably no gender bias at some institutions and in some cases). The second interpretation is that, at this stage, the findings are inconclusive: the jury is still out. But this in itself suggests that sex-based discrimination could not be hugely prevalent in STEM; if it were, it would be easier to detect a clear signal and the research would paint a more consistent picture of the situation. This, in turn, suggests that factors other than discrimination – in particular, sex differences in occupational preferences – are the main explanation for the persistence of gender gaps in STEM.