Open access publishing continues to grow in popularity globally. However, uptake varies across the spectrum of disciplines. The highest levels of open access are reported in the medical sciences, closely followed by natural and technical sciences: physics, mathematics, information technology and astronomy were early pioneers of open access, while uptake in biology increased in the early 2000s. The social sciences follow in terms of prevalence of open access publishing, with humanities, law, chemistry and engineering currently showing the lowest prevalence.

Understanding publishing trends across disciplines

The penetration of open access highly depends on author behaviours, attitudes and awareness; these vary by discipline but also based on one’s location and affiliation. The following key considerations can help explain variation in open access uptake across the board:

• National and funder policies strongly affect uptake: disciplines where policymakers have been actively promoting open access show higher rates of open access publishing (e.g. medical sciences).
• National and funder policies strongly affect the open access model that authors choose when publishing, with a clear preference for the gold open access model when article processing charges are eligible for support by research funders.
• In some disciplines, open access journals are sometimes associated with lower quality. This misconception has been shown to slow down the penetration of open access in disciplines such as chemistry, engineering and the social sciences. These barriers should be considered when seeking to launch or promote a journal in these areas.

New journals may also have to consider their disciplinary niche when it comes to forming an editorial board. The considerable growth in multidisciplinary journals and megajournals means that editorial boards should be sufficiently diverse to cover the range of subjects that authors may focus on.

Implications for author and peer reviewer guidelines

A new journal should not base its author guidelines on any existing examples or generic templates without considering the disciplinary differences of their field. Based on the discipline(s) served by a journal, it will be essential to ensure that guidelines for authors and peer reviewers cover appropriate reporting standards (e.g. EQUATOR – medical sciences, ARRIVE – animal research), reproducibility (e.g American Economic Association – economics, Executable Research Articles – life sciences) and expectations around research ethics and integrity. Additionally, journals may decide to require the deposit of specific research objects via data repositories at the time of submission or publication (including for peer review purposes). The research objects within the scope of a journal’s policy will need to closely mirror disciplinary practices and may include for example research data and research code/software.

Finally, preprint posting is common in physics, mathematics, astronomy, and information technology (e.g. arXiv) and in the social sciences (e.g. SSRN). Preprint posting has also been growing in popularity in biology (e.g. bioRxiv), medical sciences (e.g. medRxiv) and in chemistry (e.g. ChemRxiv). Preprint posting should be covered by journals operating in disciplines where this is common, as part of author guidelines.

Ideally, the above observations on guidelines and preprint posting should also be reflected in the journal’s submission system according to the target discipline(s), to ensure that relevant information can be captured early on in the process and can then be included as part of the article’s metadata at the time of publication (including via Crossref).