In January 2024, we compiled the scholarly output published between 2018 and 2022 by authors affiliated with ten Catalan universities from the Web of Science (SCI, SSCI, AHCI and ESCI). The search query, limited to “articles” and “review articles”, was formulated using the authority names employed in the Web of Science affiliation index. This strategy led to a mix of languages in the affiliation names, as the Web of Science uses the English name for some universities and the Catalan name for others. Although the index may not capture all possible names for a university, the number of records retrieved suggests that the search provides a representative sample of the scientific production of Catalan universities.

We retrieved 95,041 records. To ascertain the share of gold (33.0%) and hybrid (13.4%) OA outputs, we relied on the information sourced from OurResearch’s Unpaywall database through the Web of Science.⁴ To discern the fees charged by gold OA journals, we used DOAJ data⁵ downloaded on 29 January 2024. Exchange rates were calculated using the Google Finance currency converter.⁶

Data on expenditure was compiled by the ten Catalan universities (CSUC, 2023), encompassing: a) library expenditure allocated to journal subscription licenses; b) library expenditure allocated to transformative agreements; and c) APCs paid by Catalan universities through external research grants and university funds to cover the publication of gold and hybrid OA articles.

To identify funding support for gold and hybrid OA articles, we searched the “funding text” field in the Web of Science records for any of these five expressions: “APC”, “article processing charge”, “author fee”, “open access” and “publication fee”.

Our search yielded 1,498 excerpts acknowledging funding support to pay APCs, with 634 corresponding to gold OA articles and 864 to hybrid OA articles. These excerpts correspond to approximately 4% of the articles published in gold journals with APCs (i.e., excluding articles in diamond journals) and hybrid journals. While this sample does not fully represent the entire population of OA articles and the analysis was constrained by the lack of standardization in acknowledgements, these excerpts offered valuable insights into how researchers finance OA publication fees.

For gold OA articles published in 2022 with corresponding authors affiliated to a Catalan university, we collected institutional email addresses (i.e., disregarding “gmail” and similar addresses). Corresponding authors of hybrid OA articles were omitted from the survey as hybrid OA publication is often not a deliberate decision of authors but occurs due to inclusion of journals in transformative agreements.

We identified 491 potential participants for the survey. The survey instrument was an anonymous online form tailored to elicit qualitative responses to three open-ended questions:

In March 2024, participants were invited via email to reply to the online survey. Each email was personalized to the recipient. Employing a critical incident technique, email messages referenced in the title of the article for which participants were corresponding authors and prompted them to consider the article while they were responding. Responses were collected through Microsoft Forms. On survey closure, 202 participants had responded, which was a response rate of 41%.

The raw data supporting this article is available at Borrego (2024).

Between 2018 and 2022, researchers at Catalan universities published 95,041 articles. The annual output increased regularly from 2018 to 2021. However, the number of articles published in 2022 was lower than in the two previous years (Table 1). This decrease is unlikely to be attributable to delays in the indexing process by the Web of Science, since data were not collected until January 2024 to mitigate this possibility.

Despite a slight decrease in the number of gold OA articles published in 2022, the share of gold OA escalated from 25% in 2018 to 39% in 2022. Hybrid OA increased in absolute and relative terms, to reach 21% of the output in 2022.

The 31,319 gold OA articles published during the five-year period under scrutiny were published in 2,915 journals. We located 2,503 of these journals (86%) within DOAJ. These journals had published 28,931 (92%) of the Catalan gold OA outputs. Table 2 shows the fees charged by these journals.

The results reveal that, despite 37% of gold OA journals operate on a diamond model (i.e., they do not impose APCs), most articles (82%) were published in journals that do charge APCs. In fact, 62% of the articles were published in journals that charge more than 2,000 euros per article, even though these journals comprise only 31% of the publishing outlets. A rough estimate, based on the simple sum of publication fees, suggests that collective APC expenditure for publishing the gold OA scholarly output of Catalan universities between 2018 and 2022 would amount to 59 million euros.

Hard data collected by Catalan universities included library expenditure on journal subscriptions and transformative agreements, plus APCs paid by researchers for gold and hybrid OA publication (CSUC, 2023). The data underscored a consistent upward trajectory in expenditure, rising from 10.4 million euros in 2018 to 14.7 million euros in 2022. In 2022, funds were allocated as follows: 45% for transformative agreements, 37% for journal subscriptions, and 18% for APCs.

An examination of the combined costs of journal subscriptions and transformative agreements revealed that expenditure increased from 8.68 million euros in 2018 to 12.03 million euros in 2022 (Figure 1). The stabilization of expenses from 2021 to 2022 is explained by the reduction in APC payments, which decreased from 3.10 million euros in 2021 to 2.67 million euros in 2022. This decline aligns with the decrease in the publication of gold OA articles and the influence of transformative agreements that attract to hybrid OA some articles that might otherwise have been published in gold OA journals, as outlined in Table 1.

We retrieved 1,498 excerpts acknowledging funding to pay APCs, either for gold (634 excerpts, Table 3) or hybrid (864 excerpts, Table 4) OA articles.

In gold OA articles, nearly half of the excerpts acknowledged support to cover publication fees from a non-CSUC university or research centre, i.e., authors of these articles acknowledged support from an APC funding scheme available to coauthors affiliated with a different institution. For example:

The second source of funding to pay APCs were research grants. In some instances, distinguishing between the first two categories was challenging. As a rule, we classified as “research grants” the excerpts that mentioned a specific grant, whereas general excerpts acknowledging a non-CSUC funder were classified as “non-CSUC research centres”. In certain instances, such as the first example below, the same grant encompassed both the expenses for conducting the research and the APCs. In other cases, such as the second example, authors relied on separate sources for each purpose.

About one in every ten acknowledgements recognized funding from a CSUC university. Some acknowledgements made general references to funding from a university (as in the first example below), while others specified support from a particular department (as seen in the second example) or a funding programme specifically designated to cover APCs (as illustrated in the third example). Occasionally, authors indicated that the institution only partially covered APCs without explicitly stating the source of the remaining funds (as observed in the second and fourth examples).

Alternative sources of APC funding were less common yet noteworthy. For instance, 3% of acknowledgments recognized the support from private companies, predominantly from the pharmaceutical industry.

Certain acknowledgments revealed that authors personally financed the APCs, either through fellowships (as in the first example below) or using their own resources (as illustrated in the second example). Additionally, some excerpts indicated that authors did not receive external funding to cover APCs (as shown in the third example), which implied that they incurred the expenses themselves. This finding suggests that the practice of covering APCs with personal funds might be more widespread, although authors may not necessarily feel compelled to include this information as part of an “acknowledgment”.

In a few instances, the publisher waived APCs and bore the expenses of OA publication.

Lastly, certain references to transformative agreements in acknowledgments for gold OA articles may potentially be errors, either within the acknowledgment itself or in the classification of articles as gold OA.

Regarding hybrid OA articles (Table 4), nearly two-thirds of the acknowledgements indicated that the articles had been published within the scope of transformative agreements. Most acknowledgments referenced the CRUE-CSIC transformative agreements in Spain (345 excerpts), Projekt DEAL in Germany (120 excerpts) and CRUI-CARE in Italy (68 excerpts). Other categories of acknowledgments resembled those discussed for gold OA articles. However, the low percentage of hybrid OA articles acknowledging funding from research grants to cover APCs is notable.

Tables 3 and 4 provide valuable insights into the disparities between the estimated gold APC expenditure based on publication data (totalling 59 million euros between 2018 and 2022, as indicated in Table 2) and the actual APC expenditure data collected by Catalan universities (amounting to 11.4 million euros between 2018 and 2022, as depicted in Figure 1). The factual APC expenditure data gathered by Catalan universities solely pertains to APC funding programmes managed by Catalan universities themselves and APCs funded by research grants managed by Catalan universities. The remaining sources of funding identified in Tables 3 and 4, including APC funding schemes and grants managed by non-CSUC universities and research centres, contributions from private companies, authors’ personal assets, and waivers by publishers, would be allocated to cover the costs associated with the remaining gold OA articles. Furthermore, the results shed light on the frequent utilization of transformative agreements drawn up by coauthors’ institutions for hybrid OA publication.

Two hundred and two corresponding authors of gold OA articles affiliated with Catalan universities participated in the open-ended online questionnaire. When respondents were asked why they submitted their manuscript to a gold OA journal, they typically gave a mix of reasons.

By a significant margin, the most common rationale cited was the journal’s impact factor, which was mentioned by 52% of the respondents. The OA availability of the article emerged as the second most prevalent reason, noted by 38%. In many instances, respondents indicated that they valued the visibility afforded by OA availability, although a handful justified their decision to publish in OA journals based on funders’ requirements.

Other criteria were also frequently mentioned, albeit to a lesser extent. These included the speed of the publication process (30%), the prestige of the journal (25%), its scope (20%) and its target audience (18%).

Some respondents relied on their specific knowledge of the journal: they had been invited to submit a manuscript (7%), were responding to a call for papers (4%) or were members of the editorial board (3%). In an additional 3% of instances, the authors had not initially submitted their paper to the publishing journal, but rather to another journal that rejected it. They were then offered the opportunity to submit the paper to the current journal as part of a cascade submission process.

When asked about the resources they had utilized to cover the APCs, nearly half the respondents (49%) indicated that they had used research grants. In addition, authors relied on university funding schemes (21%). Another comparable percentage (19%) depended on their “own” funds, a miscellaneous category encompassing research funds resulting from contracts with private companies and, in some cases, authors who stated that they had directly paid the APCs from their personal funds. All these categories were not mutually exclusive, as respondents often combined funding sources. A notable proportion of respondents (17%) stated that they had not paid APCs because they had been invited to submit, were members of editorial boards or had vouchers given in exchange for reviewing.

In the section where authors were invited to provide additional comments regarding OA publication and associated APCs, several recurring themes emerged. Foremost among these was criticism of the high fees charged by gold OA publishers in comparison to the perceived value of the services offered. Many respondents described publication fees as “abusive”, “excessive”, “exorbitant”, “scandalous” or “shameful”. One respondent estimated that 20% of their research funding was allocated to APC payments, while another expressed dissatisfaction with having paid 10,000 euros for the publication of a single article in a “prestigious” journal. Another respondent expressed “remorse” for utilizing public funds to publish their research results.

Despite this criticism, authors acknowledged the pressures of the publish-or-perish model, which compels them to publish as much as possible. In the words of one participant “researchers need these journals to advance in their professional careers and journals take advantage of these needs to impose abusive fees”.

Some gold OA journals were praised for offering desirable features such as high impact factors and speedy review and publication processes. However, several respondents lamented a misalignment between funder requirements for OA publication and the criteria used in research assessment exercises.

Additionally, some respondents called for universities to allocate more resources to cover APCs, citing the scarcity of funding for this purpose. They suggested implementing “transformative agreements” like those used for hybrid OA to cover gold OA fees.

Furthermore, authors raised concerns about inequity, noting that only those with resources can afford to publish in gold OA journals. This creates a situation where information may not be published due to authors lacking the resources to pay APCs, thus exacerbating disparities in access to scientific knowledge. As one respondent put it, “equality of opportunities has evaporated, and with it, science suffers”.

Several respondents also mentioned perceived conflicts of interest among gold OA publishers, who may be incentivized to publish extensively to increase revenue. This has led to scepticism about the prestige of gold OA journals compared to traditional ones. Some authors expressed concerns about the rapid review process and perceived lack of quality in gold OA journals that prompted them to reconsider their publication strategies. In the words of one respondent, “paying to publish makes a bad impression”.

Despite the criticism, few respondents offered alternative solutions. Only a couple suggested the possibility of funding peer-reviewed repositories. Diamond journals were hardly mentioned.