Despite Progress, Disparities Persist in Science and Peer-Review Publishing

In the summer of 2020, when I was still the vice president of publishing for the American Gastroenterological Association (AGA), our editorial staff and editors were moved to reflect, as many in our industry were, on how we might rise to the call of ensuring our family of journals was diverse, equitable, and inclusive. On the face of it, we could see that our boards of editors were lacking diversity—but that observation was without data (we didn’t collect demographic information about our editors). Furthermore, we didn’t systematically collect such data on our authors or reviewers, and so while we had a sense that stark disparities and inequities likely existed within our portfolio, we lacked the necessary information to support that thought. It was then that we realized initiating change would be a complex and multi-layered effort. And for me personally, I took a step back and began to educate myself on the scale of these issues and their impact on the entire research ecosystem. In this post I will highlight some of the research that has informed my own thinking on matters of DEI in scholarly publishing, as well as share highlights of a guidance document from the Society Publishers’ Coalition on initiatives journals might consider as they embark on bringing about much-needed change. To be clear, this post focuses on diversity in the context of gender and race/ethnicity—many other forms of diversity exist, including cultural, religious, age, sexual orientation, disability, among others.

The lack of diversity, equity, and inclusion in science has been well established. For example, while women researchers have made considerable gains over the last several decades, gender inequity stubbornly persists. A study published in the journal Quantitative Science Studies analyzed gender disparities in more than 140 of the world’s top international research awards, finding that women’s share of these awards represented an annual average of just 19% during 2016-2021. Moreover, in an article published in PLOS Biology, Holman et al. showed that despite recent progress, the gender gap in science will likely endure for generations, especially in the fields of surgery, computer science, physics, and math. And as is the case in other sectors, women scientists with PhDs in North America earn substantially less than men in similar roles, according to a 2021 survey conducted by the U.S. National Science Foundation. 

This persistent underrepresentation and lack of diversity isn’t limited to gender. Racial and ethnic disparities continue to plague science, with only modest gains being made. According to a 2021 Pew Research Center report that analyzed nearly 20 million U.S. scientists, Black researchers only accounted for 9% of the STEM workforce in 2019, whereas Hispanics represented 8% of the workforce (Blacks make up 11% of all employed adults whereas Hispanics make up 17% of all employed adults). Furthermore, an analysis of grants awarded by Cancer Research UK found that members of racially and ethnically underrepresented groups were less likely to receive funding. Other research has shown that while underrepresented racial groups produce higher rates of scientific novelty, the scientific community devalues and discounts these contributions. In the study cited here, the authors considered “impactful novelty,” which they measured “how often a thesis’s new conceptual linkages are adopted in ensuring documents each year.”

Data showing the underrepresentation and lack of equity, as well as bias, in the publishing component of the research ecosystem isn’t as abundant. This is despite those of us in the scholarly publishing industry having observed such disparities for years. The lack of data is, in part, because associations and publishers haven’t systematically collected this information as well as because of privacy concerns related to GDPR. Complicating matters is the international nature of journals, as there is no universally applicable framework for measuring race and ethnicity that makes sense for researchers from diverse geographical and cultural backgrounds.

In the absence of comprehensive data, some researchers have employed computational algorithms to assess diversity. Although these methods, which estimate demographic information based on names, are far from perfect, they can be useful if best practices are followed to mitigate the pitfalls that come with this approach. Some of this research has shed light on biases in the peer review process. For instance, a study conducted by the Royal Society of Chemistry, analyzing more than 700,000 manuscripts submitted to its journals from January 2014 through July 2018, revealed bias against women at every stage of the publishing process. Other investigations have delved into citation patterns, revealing a tendency to under cite authors of color relative to their representation. Additionally, studies have shown that research from historically underrepresented groups tends to be overrepresented in less-cited topics and receives fewer citations within those topics. And a 2023 study in PNAS, also using race- and ethnicity-detecting algorithms, found that historically underrepresented ethnic groups had to wait for longer periods of time from the point of submission to the time of acceptance. 

Several prominent organizations, such as the American Association for the Advancement of Science, the American Chemical Society, and the Royal Society, have publicly disclosed the demographic composition of their authors and reviewers, revealing a makeup of predominantly White researchers. This pattern is not uncommon among editorial boards as well. According to an article published in Nature in February 2022, editorial boards and professional staff of several high-profile journal families such as Science, the JAMA Network, and Nature are predominantly composed of White editors (given that this article is now more than a year old, perhaps the composition of these boards has made progress toward diversification). Moreover, editors-in-chief often tend to select board members primarily from their own geographic regions, further perpetuating the lack of diversity.

As we know, large-scale efforts are now underway to address the issue of diversity in scholarly journals. The Royal Society of Chemistry, for example, published a framework aimed at enhancing diversity and reducing bias. They plan to collect and analyze demographic data concerning authors, editors, and reviewers, as well as the authorship status and outcomes of peer review. Based on this analysis, they will establish benchmarks and success metrics to track progress. Notably, the Joint Commitment for Action on Inclusion and Diversity in Publishing, comprising over 50 publishers, representing more than 15,000 journals, set out to start collecting race, ethnicity, and gender information of their authors and reviewers from 2023 onward. This initiative represents the largest-scale effort to date in understanding the representation within journals and investigating the existence of bias in the peer review process.

The Problem is Daunting—What Can Your Journal Do to Influence Change?
But what can you do, at your individual journal (or journals) to initiate change? When the journal editors and staff were pondering this question at AGA, I was admittedly overwhelmed. The task was incredibly daunting and, at that point, journals in our space (biomedicine) were just starting to consider what needed to be done; hence, there weren’t many models to follow. We knew we had to act and yet we felt paralyzed at the same time, not certain what issue to tackle first or how. We grappled with questions like “What can we do that would have the most impact?”, “How do we know how big of a problem we have with our journals?”, and “How do we demonstrate progress?” There were many, many additional questions we mulled over. I often thought to myself, if there was just a starting point, a set of recommendations perhaps, that we could consider for our journals, that would at least point us in the right direction.

Fortunately, such a guidance document now exists: “Recommendations for Equity, Diversity, and Inclusion Initiatives for Society Publishers”. This document from the Society Publishers’ Coalition (SocPC), written by a multi-author team of SocPC members, endeavors to provide guidance to society publishers on how to address matters of equity, diversity, and inclusion within their journal programs (the use of EDI vs. DEI in this document is reflective of the former term being more commonly used in the U.K., where SocPC originated.) SocPC is an organization comprising learned societies, representing a diverse range of academic disciplines who are addressing the demand for open scholarship and share the common goal to see “an orderly and sustainable transition to open scholarship and to improve the efficiency of the scholarly communication ecosystem for the benefit of researchers and society at large in a fair and sustainable way.”

Figure 1: Recommendations document from the Society Publishers’ Coalition on DEI efforts scholarly journals should consider pursuing

Over the course of several months, the writing group, of which I was a member, discussed the various ways in which journals are approaching matters of DEI, organizing the issues into six categories: Policies & Statements, Training/Education/Awareness, Peer Review, Processes, People, and Open Scholarship.

Our intention for this document is to serve as a collection of ideas journals can pursue—a one-stop-shop, if you will. It’s by no means exhaustive and it represents the opinions of the authors—additionally, it’s reflective of the writers’ locales (North America and Western Europe). We are fully aware that what we suggest might not necessarily apply to all scenarios in all parts of the world.

The initiatives included in the guide range in their complexity of implementation, with the goal of meeting journals “where they are” in their DEI journeys. For example, for journals just starting to explore how they can make changes, they might consider adopting an inclusive-language policy or taking stock of the demographics of their editorial boards. For journals that are already engaged in DEI efforts, they might think about more complex efforts such as reviewer training around implicit bias, collecting demographic data of their authors or reviewers, or developing a regularly recurring DEI column that aims to raise awareness about DEI in the journal’s field, showcase research on DEI in that field, or educate the journal’s community. Wherever your journal is in its path toward greater diversity, equity, and inclusion, the recommendations guide will have something for you and your editors to consider.

With so much work to be done to level the playing field in scholarly publishing, it can be overwhelming to know where to start or how to advance your journal’s already existing program. Our hope with this guide is that you’ll find inspiration within its pages to either start the discussion with your editors or to think about what effort you’ll pursue next. We welcome your thoughts on the guide and are interested, too, to hear what DEI efforts your journals are engaged in.