Blog Series: Patterns in Open Access Publishing - Pt. 3

Observations:

The first two posts in this series (here and here) reported the results of an analysis of 257 open access articles published in 174 different journals that had been referenced on my blog over a 5 1/2 month period between 1 Oct. 2019 and 15 March 2020. This post shares a few observations and takeaways regarding patterns of open access (OA) publishing in the forced migration context.

1. How do forced migration authors make their articles open access?

The articles in my sample were distributed across four different OA categories: gold, hybrid, diamond and green. The first two OA types charge authors publication fees (APCs), while the latter two are cost-free to authors. In my sample, the scale was tipped slightly in favor of APC-articles, with 53% of the articles published in gold and hybrid journals versus 47% published in diamond journals or deposited as eprints via the green route.

Even after the articles were divided into two author subsets -- Global North and Global South -- the pattern did not change very much. There continued to be a similar tilt toward publishing in APC journals for the GN group: 54% gold and hybrid OA versus 46% diamond and green OA. And the GS group had essentially an even split of articles published in APC versus APC-free journals.

Where the picture looks different is within the Global South author subset: 76% of articles by Southern authors alone (i.e., solo authors or South-South co-authors) were published in the diamond category, whereas 77% of the articles with a collaboration of North-South co-authors were published in the gold and hybrid categories.

A larger study is necessary to determine whether or not this pattern exists beyond my small sample. I plan to conduct an analysis of refugee-related articles indexed in the Directory of Open Access Journals (DOAJ), which will include a mix of both diamond and gold categories.

2. What "pay-to-publish" costs, if any, are required to make forced migration articles open access?

As noted above, just over half of the open access articles in my sample were published in gold and hybrid journals that charge an APC when an article has been accepted for publication. An analysis of the fees charged by the journals in my sample found that hybrid OA APCs were significantly higher: an average of $2995/median of $3000 versus an average of $1646/median of $1595 for gold OA APCs. This is consistent with findings by other studies (see, e.g., Pinfield et al. 2017).

Who actually pays these APCs? While the bill may go to authors, the fee is often covered by an author's academic institution, research grants, or agreements between universities and journal publishers (see, e.g., the participants in the Public Library of Science's Institutional Account Program). Or the APC may be discounted through other funding arrangements such as membership programs with specific journal publishers (see, for example, BioMed Central's membership list). Finally, some authors, primarily those based in low-income countries, may be eligible for APC waivers.

Since the APCs for hybrid OA are generally higher, these fees are often excluded from funds that cover publication costs. As an alternative, certain types of agreements have been negotiated between publishers and university libraries that cover hybrid publishing costs for authors. (See examples of  these agreements specifically in the Netherlands or check this registry for a more comprehensive listing.)

It was not possible to verify whether or not the authors in my study paid an APC, or received a discount or waiver. In general, though, it would appear that authors based in the Global North have access to a greater variety of funding support options than authors based in the Global South. A quick look at the various lists linked to above shows that the entities funding APCs or entering into agreements with publishers are largely based in Western Europe and North America.

Therefore, finding that the GS authors in my sample 1) tended to publish in diamond (APC-free) journals, 2) collaborated with GN authors to publish in gold journals, and 3) were less likely to publish in hybrid OA journals, is likely consistent with the results of other similar analyses.

3. What about the green open access category?

Only 15% of the articles in my sample fell into the green OA category. Within each of the two subsets, 19% of articles by GN authors were green, while a mere 3% of articles by GS authors were green. While this option offers the greatest flexibility to authors in terms of publishing in their journal of choice, it also requires that they invest more time and effort in the process, such as keeping track of eprint versions of articles, understanding publisher policies regarding self-archiving (particularly embargo periods for postprints), and learning where eprints can be deposited. A better understanding of author attitudes towards green OA is required in order to draw broader conclusions.

That said, in order to deposit an eprint at all, an author's article must first have been published in their journal of choice. To what extent is this the case for GS authors in the forced migration context? Two recent analyses found that articles published in the Journal of Refugee Studies and Migration Studies were more likely to be written by authors affiliated with institutions in the Global North than by authors affiliated with Global South institutions. More research is required to determine if this pattern persists across other key migration journals.

4. What is the geographic representation of forced migration authors who publish open access articles?

In my sample, a majority of the open access articles (75%) were produced by authors affiliated primarily with institutions in Europe, North America and Australia. Only 25% of the articles had at least one author affiliated with a Southern-based institution and only 18% listed a first or solo author from the Global South. Overall, the 719 authors were affiliated with institutions based in 57 different countries, 29 in the North and 28 in the South.

Since my study sample was comprised of articles that had been referenced on my blog, it is limited in a few key ways. First, in terms of language: Most articles that I reference are in English, with a few others in French or Spanish. Second, in terms of subject area: For example, I dramatically reduced my coverage of health-related issues last year. Greater linguistic and thematic diversity would undoubtedly have increased Southern representation.

To date, a greater number of Northern institutions have adopted open access mandates that require researchers to provide open access to their research. This may help to explain the higher percentage of GN authors represented in my sample. Yet, as this article points out, the share of open access articles being published in Asia, Africa and Latin America is growing quickly. As such, employing new strategies to retrieve relevant open access articles from other types of journal databases and portals is essential to ensure the ongoing development of a more diverse collection.

Related posts:
- Blog Series: Patterns in Open Access Publishing - Pt. 1 (21 April 2020)
- Blog Series: Patterns in Open Access Publishing - Pt. 2 (23 April 2020)