It is well known that women are under-represented in senior science and research positions. This is true even in Denmark, which has long been considered one of the most advanced societies when it comes to gender equality. Although stories of sexism in science often focus on explicit bias, more-subtle factors are widely influential too.

Universities like to think of themselves as meritocracies. Indeed, one of the arguments used against programmes that aim to proactively promote the careers of women scientists is that scientists must be recruited on talent alone. When criticized over the appointment of (another) male scientist to a senior role, universities often respond by pointing to rules and policies about how vacancies invite all to apply.

I carried out an analysis that raises some troubling questions about how closely universities follow these principles (M. W. NielsenSci.PublicPolicyhttp://doi.org/7q6;2015). In the decade to 2013, about one-fifth of associate- and full-professor positions at Aarhus University, one of the largest in Denmark, were filled through a ‘closed’ recruitment procedure: no advertisement and usually just a single applicant. The share of female candidates for such positions is particularly low — just 12% of applicants for full professorships were women.

With ‘open’ recruitment, the proportion of female applicants for full-professor roles rises to 23%. But a significant proportion still attracted only a single applicant, suggesting that the adverts were being written to target a specific candidate, or timed to fit their career progression. Evidence suggests that similar practices, to various extents, are common at other Danish research institutions and abroad.

Despite institutional efforts to make recruitment more robust and transparent, and a 2008 Danish ministerial decree that “professorships must be advertised internationally, except under special circumstances”, my analysis shows that the use of closed recruitment at Aarhus University increased from 8% of tenured appointments in 2004–08 to 30% in 2009–13.

Such appointments do not usually break any rules; loopholes can be exploited in most cases. But the numbers suggest a lack of institutional commitment to overarching organizational and legal stipulations. And they confirm what most academics may already suspect: rising in the ranks is not a question merely of what you know, but of who you know.

This puts women at a particular disadvantage. Academic advancement through back-door hiring largely depends on reputation and visibility to the local gatekeepers, and women lose out under such procedures for two reasons.

Disadvantages accumulate over time through self-reinforcing effects.

First, women have been shown to have weaker personal ties to the core of the concentric circles of academic networks, making them less visible to decision-makers. Second, scholars have argued that male decision-makers’ desire for organizational certainty and their attraction to candidates with whom they share values and behaviour, create subtle and often unconscious practices of ‘male closeness’ and ‘gender homophily’ (preference for someone similar to oneself).

Gender scholars have previously shown that discrimination may be particularly prevalent in organizations that pride themselves on being meritocratic. Strong institutionalized beliefs in meritocracy are more likely to discourage people from paying attention to their own implicit biases and prejudices.

Sure enough, department heads seem unaware of the implications. I interviewed 24 at Aarhus about whether and how issues of gender influence their recruitment and selection practices. I was told frequently that “gender doesn’t play any part”, “for us it’s all about getting the best” and “what we look at is quality”.

As Liisa Husu, a gender-studies researcher at Örebro University in Sweden, has pointed out, the myriad disadvantages facing female academics often operate as “non-events” (Nature 495, 35–38; 2013). Women are not being taken into account, encouraged or asked along to the same extent as their male colleagues. Seen as separate occurrences, such non-events may seem harmless. But just as academic success will often accrue to the already successful, marginal disadvantages accumulate over time through self-reinforcing effects, with clear implications for gender stratification.

With an interest in addressing the gender-equality challenge, Aarhus University provided the data for this study. It acknowledges my findings, and is currently working on a plan to improve the situation, which should be announced soon.

People who use the word meritocracy as a positive depiction of society are probably unaware of its original satirical and pejorative connotations. It works only if everyone has the opportunity to compete.

If we really believe in meritocracy as the main principle for sorting academics into positions, we must become better at focusing on the subtle and unconscious gender biases enmeshed in our day-to-day work activities. At stake is not just women’s participation in science, but also the stature, integrity and legitimacy of a scientific system renowned for its conformity to the meritocratic ideal. Although all researchers in principle should be considered equal, some remain more equal than others.