Gender-equality Paradox in STEM
One of the major transformations that shook the workplace in the 20th century, including in Armenia, was the seemingly unprecedented integration of women in many areas of the economy, working together with men in almost all fields and occupations. However, while some previously male-dominated professions now have large shares of women (such as accounting), the representation of women in STEM (Science, Technology, Engineering, and Mathematics) remains very low.
Many governments and organisations attribute this disparity to harmful stereotypes and biases against women in STEM. To address this issue, various programs, even targeted scholarships or quotas, have been implemented to promote equal representation. Scandinavian countries have made significant progress in this area, implementing a lot of programs to promote and mainstream women in STEM.
The surprise for many social scientists came in the early 2000s, when empirical data showed that the Scandinavian countries have some of the lowest shares of women in STEM and men in humanities. In fact, the more developed the country, as measured by a gender equality index or economic productivity, the greater the differences in occupational choices in terms of which fields men and women go to – what researchers have called “The Gender Equality Paradox”. Armenia is no exception, having had a higher share of STEM graduates being female (40% in 2018) than most European countries. Over the last two decades, more research has been conducted on this topic, and while the empirical findings are largely accepted, different theories have emerged to explain the underlying factors at play.
What Explains This Phenomenon?
Two theories have been put forward – broadly categorised as the “biological” and the “social” one.
On the face of it, growing differences in occupational choices may support the idea that these differences have biological origins. If we assume that individual interests and preferences are influenced by both social and biological factors, then in countries where social factors are relatively reduced, the role of biological or innate factors can become more prominent. As researchers have argued, some women in developing countries may pursue more technical careers because they need a high salary for a decent quality of life, and not always out of interest. On the other hand, in developed countries, there are more opportunities to earn a good salary without entering technical fields. As a result, those women who may not be as interested in technical occupations no longer feel pressured to pursue STEM fields.
It has also been pointed out that differences in other commonly used personality metrics are also larger in more gender-egalitarian societies. They note that growing disparity in occupational choices is not an isolated phenomena, but part of a larger trend. The personality metric that is posited to explain difference in STEM interests is “people-things orientation”, i.e. whether someone is more interested in working with people or with objects. On average, women display a higher interest in working with people, which ends up being humanities or life sciences (such as biology and medicine), while men display a higher interest in working with things (e.g. engineering and physics). Not only are these personality differences consistently present across cultures, but they also tend to grow as societies become more gender egalitarian, which has led part of the scholars to infer that these differences (including in people-things orientation) are to some degree innate (Lippa et al, 2008; Schmitt et al, 2016; Stewart-Williams and Halsey, 2021).
While some scholars have seen evidence of biological factors, others have argued that it is still certain stereotypes and biases driving these disparities in occupational choices. As an example, society tells us from early on that men and women are different; while in traditional societies that difference was manifested in the “working father – stay-at-home mother” formula, in more developed economies everyone works, so the difference we project onto ourselves is which fields are feminine and which are masculine. For instance, some studies have shown that individuals in more developed countries are more likely (whether consciously or unconsciously) to associate humanities with females, and technical sciences with males. Another study estimated the presence of stereotypes by looking at how high school girls and boys answer differently to two statements – “doing well in maths is completely up to me” and “my parents think that maths is important for my career.” Showing that boys answer more positively to those statements than girls do, and that the difference in answers increases in more gender-egalitarian countries, the authors argue that this is evidence of a stereotype-driven phenomenon.
In my view, the main problem that these studies run into is “the chicken or the egg” issue. It is not as easy to differentiate mere associations from actual biases, and the question is to what extent it is stereotypes that drive women out of STEM, as opposed to women going predominantly into other fields that generate the stereotypes or associations as a result. While I find the biological hypothesis more likely at this stage, it would be premature to say that the debate is closed, and a healthy dose of agnosticism would be advised. At any rate, we probably do not really understand what feature or gene (if it is in the genes after all) influences our innate interest in occupational choice, and how that could be reflected in male-female differences. Certainly, biological factors and external conditions vary from person to person, leaving some of us with more freedom than others to pursue our innate interests and talents.
Policy Implications for Armenia (and the World)
As is often the case, the more nuanced conclusions that may be drawn from academic research are contrasted with the more simplistic assumptions when it comes to policy and advocacy. Statements like “women in Armenia are only X% of tech workers or engineers”, often combined with some examples of stereotypes and life constraints, are given to imply that in an ideal world without biases we would have a close to equal representation of men and women across occupational fields and hierarchies. Such logic can be found in both individual statements (e.g. Roza Melkumyan, 2022; Hranoush Dermoyan, 2023) and those of international organisations (e.g. World Bank, 2017; Asian Development Bank, 2019; UN Women, 2021; EU4Environment, 2021). The problem with these is not their recognition that biases exist, nor their motivation to provide broader opportunities and reduce the effect of those biases. It is the assumption of knowing what the outcomes should look like in the absence of biases and stereotypes, and often the willingness to build policy on such assumptions. Furthermore, usually the mark is set at equal representation (50/50), implying absolutely no innate difference in interest – a strong assumption with often little to no evidence being asked or given. Quotas, diversity representation requirements, or incentive programs (such as targeted scholarships) would be some examples of policies aiming at a predetermined outcome.
Many such policies directly aim to achieve equality of economic and social outcomes (be it wages, political engagement, or occupational choices) across various groups, at the cost of inequality of power that inevitably arises by intervening in other people’s decisions through policy instruments, whether the desired results are achieved or not. The proponents of these policies may argue that such power is necessary to counterbalance the effect of biases and stereotypes throughout society, and that this is prerequisite for what would be considered “true” equality of opportunity. It is undoubtedly the case that biases and stereotypes will always be present throughout society. However, it is hard to determine the extent to which they are actually responsible for differences in outcomes, and therefore what the appropriate countermeasure should be. Furthermore, the influence of these stereotypes and biases is dispersed and varies across individuals and institutions, and more merit-based approaches will tend to take over insofar as they give competitive advantage. In contrast, government power is more coercive and concentrated in the hands of few politicians and policy-makers, who not only have their own biases and presuppositions, but usually pay little price for being wrong when intervening in third-party decision-making.
While Armenia currently has a higher proportion of women among STEM enrollment compared to many developed countries, it may well change. Firstly, as the country becomes more developed, given the trends found globally, the disparities in occupational choices are likely to become larger. Of course, Armenia could be an exception for one reason or another, but this is likely, given the overall picture. Secondly, the parliament passed a law that defers compulsory military service for students who enrol in STEM fields in top 5 Armenian universities. This could drive more males to apply for these technical degrees even if they may not be as interested in the subject.
In this context, attempts to implement government programs that aim at some predetermined proportion of men and women in technical fields risk growing more and more inefficient over time, as the situation increasingly diverges from the unspoken assumptions behind such “equal representation” policies. The policy recommendation for Armenia, and for any country, is to focus on broadening opportunities for individuals to discover and invest in their talents and interests, particularly in the formative years. School programs that allow students to engage with a variety of more specific and niche subjects, or career fests where they get to simulate what a particular profession’s day would look like, – such approaches can provide a chance to try something new for a girl who has been kept out of garage, or a boy who has been told that cooking is not a masculine task. Retraining opportunities for women after maternity leave may also help retain valuable talent in fast-changing tech industries. The specific details of such programs, and their suitability for public versus private domains, are questions to be discussed. Such programs will not lead to a perfect society, – no program will, – but they can provide a better outcome. Importantly, they will do so without third parties incentivizing or requiring people to hire, or pursue a particular career, on the basis of one’s gender, simply because some good-willed politicians and policy advocates decided that there’s not enough women in places where they believe more women ought to be.
