Gender at the LJIL

1. ‘Is there a Problem? What is it?’

Those were the questions raised in our June 2021 Board meeting. At that point in time, our Editorial Board had been – like the rest of the world – grappling with more than a year of the COVID-19 pandemic. Feeling the effects of being suddenly forced to work from home and struggling with care-giving duties, we were concerned with the effects of the pandemic on LJIL authors, especially women.¹ In particular, a few of us had the impression from their daily editorial work that there had been a drop in the number of submissions made by women. As an academic journal that aims to provide a platform for new and diverse voices as a way of contributing to the development of international legal debate,² we hoped that this perceived decline in women’s submissions would prove illusory, or if not that, temporary; and also that our processes and peer-review practices were not only effective in selecting excellent articles, but also fair and non-discriminatory. We therefore set out to determine empirically whether our suspicions were real by going through LJIL data.

The first step was to establish data on gender distribution of the LJIL authorship and how our peer-review and decision-making processes could be improved to foster a more diverse publication. In this editorial, I present, on behalf of the LJIL Board, the results of what was initially envisioned as an internal study on the presence of women in LJIL. Our main goal was to determine the share of men and women in LJIL submissions and published articles over time. Collecting the data served a twofold purpose: (i) data on publications shows us LJIL’s current state of affairs in terms of gender representation while (ii) data on submissions help us identify how our current processes for submission and peer-review shape this state of affairs.

Before explaining the study and our conclusions, we must draw attention to two points. The first relates to limitations in our study. The results presented below are based on a binary division of gender between men and women that is outdated. Although we do not subscribe to this binary, the limited resources and available information unfortunately did not allow us to ascertain further than the binomial ‘men-women’. Those same limitations have also prevented us from examining the effects of intersectional markers such as ethnicity, race, religion, and other identity markers. We apologize to authors and readers that do not see themselves fully represented in our results. Nevertheless, we wish to highlight that this project is just a starting point for a broader conversation and initiatives on gender representation and diversity in the LJIL.

The second point is that, for the purposes of this study, we are considering that ‘equal gender representation’ means equal presence of both men and women in our publication. Other studies on gender diversity in academic publishing have considered that it would be unrealistic to expect women to account for 50 per cent of publications in markets where they represent a much smaller share of faculty.³ We acknowledge that setting a 50 per cent threshold as a reference for gender equality may indeed set high expectations towards our own publication platform in a market that is still dominated by men, especially at the leadership levels.⁴ However, we believe that such expectations are compatible with our aspiration of contributing to an inclusive, diverse, and ever-changing scholarly community.⁵

Although based on limited data (Section 2), the results of the study are an enlightening introspective view of LJIL submissions and publications. Our data suggests that the LJIL has always had a gender gap, both in submissions (Section 3) and publications (Section 4). Although this gap has shrunken in recent years, the position of women in our journal is still unequal and has been negatively impacted further by the COVID-19 pandemic (Section 5).

2. Data collection

LJIL was not particularly well prepared for an inward study like this. Since we aim at evaluating articles on their own merits, we had not collected any information on our authors other than what was strictly necessary for contacting them and identifying any conflicts of interest (names, emails, affiliation, funding, etc.). We never collected authors’ personal information, including their gender identity, nationality, race, or ethnic background. Thus, for the purposes of this study, we needed to collect further information on the LJIL authors.

The process of collecting author data was further complicated by archival challenges. Created in the 1980s, the LJIL has, over the course of its existence, adapted its operations to the technologies available. Until 1994, authors who wished to publish their work would have to mail anonymous printed copies of their articles. In 1995, we also started to receive submissions by e-mail. At that time, our archives were organized and processed manually and further information on submissions could not be saved due to limited storage. Only from 2012 onwards – after we moved our entire peer-review system to the Scholar One platform – were we able to store information on submissions.

We therefore have two separate datasets as a starting point for this project. The first dataset covers submissions made through our Scholar One portal from January 2012 to December 2021. The information contained in this dataset includes submission dates, manuscript title and author name and affiliation, peer-review information and decisions issued, as well as their dates. This dataset is organized by year of submission and allows us to determine how many men and women submit their work to LJIL and how they perform in our peer-review system (i.e., acceptance and rejection rates). The second dataset contains only information on the articles that were accepted and published by us since its first issue in 1988 to the end of 2012, all of which are also available on the LJIL website. The data on publications is organized by year of publication, which is not necessarily the same year in which the article was submitted. Combined, these two datasets can provide us with a comprehensive overview on the gender distribution of our authors and help us detect whether any inequalities result from LJIL’s outreach (or lack thereof) or from our internal peer-review processes.

None of these datasets initially had any information about the gender identity of the authors who submitted or published their work with us. Due to our limited resources, it would be impossible to contact each of them for the purposes of self-identification. The available alternative was to have our team collect open source information that gave indication of how our authors self-identified their gender. While most of our authors had a digital footprint with enough information to determine how they identified themselves – i.e., institutional websites or professional social media listing articles and personal information containing gender-identifiable language – there were seven authors for whom we did not have enough information to determine how they identified themselves. Information on publications by these authors has, thus, not been included in the dataset.

3. The LJIL gender distribution

3.1 In submissions

As mentioned in Section 2, our information in submissions is unfortunately limited to submissions after 2012. The information presented in this section was collected from our online submission system and is aggregated based on the year of the date of submission. Respecting our policy of confidentiality on submissions, we will not share here the specific numbers of submissions nor the specific acceptance rates. Rather, this section will focus on the percentage of women in our submissions and on the differences in men and women’s acceptance rates.

Journals can shape their pool of authors through the design of submission and review processes. Such design involves determining who is going to review submissions; the criteria to be taken into consideration during this process; and how much freedom will authors have when responding to a decision to revise and resubmit. Today, the most acclaimed review method is double-blind peer-review because it contributes to the neutrality of peer-reviewers towards authors. Double-blind peer-review cannot ensure that reviewers will be neutral towards ideas and arguments, nor that the author will be able to re-submit an article that is ready for publication. Nor that implicit biases will be adequately addressed. All production of knowledge is necessarily gendered⁶ and might affect the outcome of such research, such as the nature of arguments and their style. Even if a system of peer review such as the double-blind peer-review is in place, gender bias, together with other intersectional bias, such as ethnicity or race, class and other identity markers, can still play a role in influencing decisions by reviewers.

This being said, women have been performing, on average, better than men in our peer-review system. The share of women in submission is currently 36 per cent (Figure 1), which is slightly lower than the share they occupy in publications in the same period (38 per cent, in Figure 2 below). Following a tendency observed in other fields,⁷ articles authored by at least one woman have also performed slightly better in peer review than those authored only by men. In the ten years recorded, women’s acceptance rate has been 4 per cent higher than men’s, although, in one year, this gap was as high as 14 per cent. Articles authored exclusively by men only had a higher acceptance rate than those authored by women in three of the ten years tracked in this dataset.

Figure 1. Men and women in submissions 2012-2021 (%).

Figure 2. Men and women published by the LJIL per volume/year (%).

Following the same tendency as other journals,⁸ the LJIL received a flurry of new submissions as an effect of the COVID-19 pandemic. In 2020 alone we saw an increase of 14.5 per cent in submissions when compared to 2019 numbers. However, such growth was mostly driven by men’s submissions, which grew by 21 per cent in relation to the previous year. Women, on the other hand, submitted only 8 per cent more articles in 2020 than in 2019. At first glance these numbers seem to suggest that our authors, and men in particular, became more productive during these early years of the pandemic. However, new submissions by men did not perform well in peer-review. In 2020 and 2021 men’s acceptance rate dropped by 7 per cent, compared to 2019 while women’s acceptance rates throughout the pandemic years remained consistent with those of previous years, varying less than 2 per cent from 2019–2021. The result of such a consistent performance can be seen in our 2021 publications numbers, which show only a small decline of the proportion of women published compared to men.

3.2 In publications

In order to see how our numbers have evolved over time we have aggregated them per volume/year. Although we publish several issues per year, the administration of the LJIL is mostly based on yearly volumes – i.e., the management is done based on volumes, not issues; change of policies and formats are also implemented only from one volume to the other. We have aggregated publications in years to follow this same managerial organization.

In terms of overall numbers, only 30 per cent of the authors published by the LJIL are women. Until the end of 2021, the LJIL had published a total of 1,589 pieces, including editorials, book reviews, and scholarly reviews. These articles were authored by 1,774 authors, 1,245 of whom are men (70 per cent) against only 529 women (30 per cent). The difference in the number of published articles and the number of authors is explained by co-authored pieces: We have published 178 co-authored articles. More than half (52 per cent) of our co-authored pieces were written by men and women. Of the remainder, 42 per cent have been co-authored by men and 10 per cent have been co-authored only by women.

Our gender distribution per volume suggests that, despite the clear dominance of men in our published articles, there is also room for some cautious optimism about the future of the LJIL. Although our gender distribution is far from equal, our numbers over time show an improving landscape (Figure 2). This is especially true of the last ten volumes, the number of women published by the LJIL has shot up to an average of 38 per cent, almost doubling the average number of women published in our first ten years (21 per cent). The past three volumes (published in 2019, 2020, and 2021) had at least 40 per cent women authors. The year 2020 was the first – and so far only one – where we published the exact same number of women and men (41 of each).

One of the reasons explaining the uptick is that the professional community in which the LJIL exists and operates has changed and includes more women in recent years. Law journals cannot exist separate from the professional field in which they operate and reflect that field’s make-up (including its flaws and biases as well as their strengths). Although women still struggle to reach leadership positions in international law⁹ and in the legal profession in general,¹⁰ they have gained ground in the international law market over the past three decades. For instance, studies on the presence of women in US law schools, where the majority of international lawyers appearing before the ICJ were trained,¹¹ illustrate this growth: women today correspond to roughly half of the student body – a great improvement compared to the ‘sea of men’¹² that populated these same institutions in the mid-twentieth century. Although women still face several structural challenges in academia, including gaps in pay, evaluations, and expectations,¹³ recent decades also saw a relevant increase in the number of women researching and teaching in law schools.¹⁴

In addition to changes in the professional landscape, it is worth stressing that changes in the LJIL itself may have influenced the increase in the number of women having their work published. Since its first volume was published, the LJIL has evolved into a more professional, accessible, and transparent format that may have made it more appealing to a wider net of prospective authors – including women. Early LJIL volumes were composed of two to three issues if a special issue was published, each with six to ten articles. There was also often a short anonymous editorial. The LJIL was then edited and published by students who aimed to, with the support of private donators, promote student work and research. Created in an analogical era, there were also obstacles in broadening the scope of prospective authors and readers: submissions had to be sent by courier and our readers could only access the journal physically. These factors restricted the LJIL (and most journals active at the time) to the limited community of students and practitioners that had geographical access to it.

While our data suggest that COVID-19 may have widened our gender gap in submissions (Section 3.1, above) we can only ascertain the effects of the pandemic from the volume published in 2021. This was the volume we had where pieces were submitted, edited, and/or reviewed during the pandemic. The data for this volume shows a 10 per cent fall in the proportion of women published in the LJIL (Figure 2). The reason for this can be explained by the delayed impact of the effects of the COVID-19 pandemic on the LJIL production process – issues need to be approved and sent for production to CUP six months before the expected publication date. After articles are sent to production, authors are only allowed to do minor textual corrections to proofs. Therefore, the effects of the pandemic could not be readily seen in the early 2020 issues, as authors had already completed their work either before or in the early stages of the pandemic. For example, in the last issue published in 2020 (December), most of the articles sent for production at that time were the product of work that was completed – or at least initiated – before the pandemic and would therefore not have been impacted by it.

4. Gender (in)equality and the LJIL

Law journals are platforms where members of a professional and scientific community exchange knowledge. The reputation and impact of a journal is built on its ability to ensure that the work it publishes adds something new to its field while also being scientifically reliable. Nowadays, the acclaimed method to select articles is double-blind peer-review, since it ensures that reviewers’ opinion on articles will be based solely on their substance. Double-blind peer-review exists to ensure that bias does not affect the production of knowledge, based on the premise that good ideas can come from anywhere. This selection method ensures that no one will be rejected because of who they are, where they come from, or their professional background. As our submission data shows, women have fared well in double-blind peer-review, showing that this system may help overcome some gender biases.

In the LJIL’s early years, the gender gap was particularly abysmal, with women rarely occupying more than 20 per cent of a volume. Since then, our numbers have greatly improved, with recent years averaging around approximately 40 per cent. Women are, however, still in an unequal position in the LJIL: we have only had one volume where women and men were equally represented and external factors such as the COVID-19 pandemic seem to broaden our gender gap. It is still not clear how and whether the numbers of women publishing at the LJIL will recover anytime soon.

At LJIL we believe that we can work towards minimizing the negative, systemic factors that decrease the number of women and authors from diverse backgrounds. We are also aware that gender cannot be considered separately from other identity markers. Race, ethnicity, location, origin, sexuality, and socio-economic background might impact the submission and publication of articles in LJIL, even though we do not have data available to measure the effect these factors have in our authorship. While the LJIL does not commission pieces nor exercise any control over who submits them, we have always striven to ensure that the LJIL and our review system are welcoming to authors who escape the ‘white, western men’ pattern.

Specifically in relation to gender equality, we have a policy of not accepting symposia nor to collaborate in events or initiatives where only men are participating or where there is a gross gender disproportion. We have also sought collaborations, symposia, and series that highlight contributions from authors representing different backgrounds and points of view. In addition, our current editorial processes aim at accommodating needs and particularities of authors from diverse backgrounds during peer-review and publication. Examples of such measures are the flexibility in terms of size and format for initial submissions so that authors will not have to spend time adjusting them before even receiving initial feedback; and offering editorial support to approved articles to ensure that no submission will be rejected or have its publication postponed due to lack of familiarity with our format and style or due to difficulties with writing in English.

We believe that ensuring diversity and inclusion of authors that submit and publish their work with us is key to a field of international law that is representative and open for people with a multitude of personal and professional backgrounds. Contributing to formal equality through a review system that aims at diminishing the role that biases play in selecting, reviewing and publishing an article is unfortunately not enough. As seen in this editorial, the LJIL publications reflect structural, systemic inequalities that affect the international law field more generally. We commit to be constantly seized with this matter, and will continue to develop measures and means to achieve equal representation in our journal.