The rising cost of academic publishing is causing consternation across the research ecosystem and prompting calls in Europe for a transition to not-for-profit publishing models. Rob Johnson argues that expecting emerging not-for-profit providers to offer a cheaper solution than commercial players is a surefire way to keep them small – or make them fail.

In May 2023 the Council of the European Union noted that the costs of scholarly publishing are ‘becoming unsustainable’ and called on the European Commission and member states ‘to support policies towards a scholarly publishing model that is not-for-profit, open access and multi-format, with no costs for authors or readers’. The assumption in many quarters is that the widespread involvement of commercial actors has pushed up the cost of publishing, and that not-for-profit actors could undertake the same activities at lower costs. My  recent study for the European Commission, Scenario Modelling for Open Research Europe, suggests the truth is more complex.

Unpicking the costs of publishing

I was asked to offer guidance on the operating and financing models needed to establish Open Research Europe (ORE), the European Commission’s open access publishing platform, as a collective non-profit publishing service from 2026. A number of assumptions had to be made for this purpose, most notably that ORE would operate as an independent legal entity using outsourced service providers to deliver the majority of editorial, production and technology functions. Alternative operating models, including the hosting of ORE by an established academic or international organisation, remain possible. However, treating ORE as a standalone entity makes it a useful test case for establishing the full costs of publishing academic research. It turns out these costs are higher than many might expect.

Any publishing service provider is going to incur a range of different costs, which were grouped into five main categories for modelling purposes:

• Article production costs.
• Marketing and community engagement.
• Platform development and maintenance.
• Salaries and wages.
• Administrative overheads.

Low, medium and high cost estimates for each category were developed through desk research and discussion with a number of experts. These were then combined with three different scenarios for publication growth on ORE, which would see it achieve 1,000, 2,000 or 5,000 publications per annum by 2030. The results were captured in a bespoke costing model, developed in Microsoft Excel and made publicly available alongside the full report.

Modelling scenarios for the annual costs of ORE

While the model allows a number of different scenarios to be explored, taking a central set of assumptions indicates that ORE’s operating costs would be €1.9m per annum in 2026, with 558 articles published. As submissions increase so will ORE’s costs, rising to €4.2m per annum in 2030, when it is hoped that approximately 2,000 articles would be published. ORE’s total expenditure would be just under €16m over the five period 2026-2030, as shown in Figure 1, at an average cost of €2,400 per publication.

Fig.1: Projected operating costs for ORE 2026-2030

What is striking in this analysis is that traditional ‘publishing’ activities like copy-editing, type-setting and managing the peer review process (grouped here under the heading ‘article production costs’) account for only a minority of the overall costs.  In practice, being an academic publisher today is primarily about maintaining a technology platform, running an organisation and competing for article submissions. Technology, management, and marketing costs are all far greater than is typically assumed.

Size matters

Sensitivity analysis shows that the volume of publications has by far the most significant impact on ORE’s operating costs, both in total and per article. All other factors being equal, ORE publishing 1,000 rather than 2,000 publications per year by 2030 would reduce article production costs over five years by almost €5 million. Meanwhile, increasing the volume to 5,000 publications per year by 2030 would increase these costs by €6 million, while at the same time dramatically reducing the cost per article. Variations in other assumptions, such as the level of in-house staffing, salaries, platform expenditure, inflation or the direct cost per publication could each individually lead to operating costs increasing or decreasing by €1-2 million. Should ORE be able to recover value added tax this would enable a significant saving, estimated at €1.7m over five years – but most not-for-profit platforms (at least in Europe) will be unable to avoid these costs.

Even in the most rose-tinted scenario, where several thousand articles per annum are cranked out by a skeleton staff, with minimal spend on marketing, platform development or article production, it is difficult to see ORE’s cost per publication falling below €1,000. Plug in a more realistic set of assumptions, like those used to produce Figure 1, and the cost per publication starts out at almost €3,500 in 2026 before falling to just over €2,000 in 2030 – only marginally lower than the average APC paid in 2022.

Fig.2: Sensitivity analysis for key variables – impact on costs over five years vs. central planning scenario (2026-2030)

Subsidies or scale?

Many will be quick to point out that existing publishers can and do publish articles at much lower costs than €2,000. In the Western world, most publishers achieve this in one of two ways – subsidies or scale. Scholar-led publications are able to keep costs down through hidden subsidies in the form of volunteers’ time and institutional provision of office space, software and administrative support (see Dufour et al, 2023). Other publishers, including many learned societies, cross-subsidise their open access journals with the higher revenues still generated from subscriptions. As a rule, only the largest commercial providers, able to spread their investments in technology, management and marketing across tens or hundreds of thousands of articles, have found a way to make open access publishing profitable at less than €2,000 per article.

Exposing false economies

MIT’s recent report “Access to Science and Scholarship: Key Questions about the Future of Research Publishing” observes that, ‘for-profit publishers adopting the APC-based Gold OA model have two basic ways to increase profits: publish more articles or cut costs’. As APC-based models have become the norm, many for-profit publishers have opted to pursue both courses of action simultaneously. Yet neither the publication of more articles nor the reduction of costs are necessarily desirable goals for science.

Scientific societies, who have historically been the custodians of highly-selective journals, have largely resisted the siren’s call of increasing publication volumes, maintaining quality at the expense of a steady decline in their market share and influence, and downward pressure on their revenues. At the other end of the spectrum, born OA publishers such as MDPI, Frontiers and Hindawi have pursued rapid growth in revenues and publication volumes, but are now suffering from a loss of confidence in their editorial controls. The commercial imperative to grow article volumes has placed increased strain on authors, reviewers and editors while efforts to cut costs have contributed to the spiralling numbers of retracted articles and weakened the position of not-for-profit societies.

Subsidise to scale

The solutions to these challenges are complex and multi-faceted, but an obsession with driving down the cost per publication is not one of them. The European Council has set a high bar in its call for ‘high quality, transparent, open, trustworthy and equitable scholarly publishing’. Achieving this goal means recognising the value and costs involved in academic publishing and resourcing it accordingly. For a not-for-profit platform such as ORE, this means accepting that its cost per publication, at least for the first few years, will be closer to the article publication charges levied by commercial and learned society publishers than the cost per article quoted by a typical diamond open access journal. Upfront investment is needed to establish ORE as a not-for-profit publishing infrastructure, which must be spread across a relatively small number of publications in its early years. This approach will provide agility in the first few years of ORE’s transition into a collective publishing enterprise, allowing it to overcome the scaling and sustainability challenges that typically constrain diamond open access journals reliant on voluntary labour.

Not-for-profit publishers must be subsidised at a level sufficient for them to scale, not simply to survive, or they will remain too small to offer a viable alternative to commercial players. Scale creates the potential to reduce costs in the long term, including per article costs, but this aim must be balanced with the need to maintain quality, transparency and equity. There is no guarantee that emerging not-for-profit publishers like ORE will succeed but expecting them to offer cost savings over established commercial publishers is a surefire way to see them fail.

The content generated on this blog is for information purposes only. This Article gives the views and opinions of the authors and does not reflect the views and opinions of the Impact of Social Science blog (the blog), nor of the London School of Economics and Political Science. Please review our comments policy if you have any concerns on posting a comment below.

Image Credit: Mathieu Sterne via Unsplash.