You don’t know what you’ve got till it’s gone: the changing landscape of UK learned society publishing

Introduction

Learned societies play a pivotal role in shaping the scholarly publishing landscape, having contributed significantly to academic discourse for centuries. While learned societies engage in a multitude of activities, publishing remains a cornerstone for many societies in the United Kingdom, often serving as a vital revenue stream that underpins fulfilment of their missions. The evolution of scholarly communication, marked by the shift towards open access and the growing dominance of large commercial publishers, has called into question the future role of societies in publishing. However, the impact of these changes on the society publishing landscape is not well understood, with attempts to raise concerns about the health of society publishing often reliant on anecdotal evidence.1

This study draws on a longitudinal dataset of 277 UK learned societies covering the period 2015–2023 to provide evidence-based insights into the changing nature of society publishing. Specifically, it sets out to answer the following three research questions:

1. Have UK learned societies chosen to outsource more of their publishing activities to third parties since 2015?

2. How has the choice of outsourced publisher for UK learned societies changed since 2015?

3. How have UK learned societies’ revenues from publishing changed since 2015?

By exploring these questions, this study aims to contribute insights into the adaptive strategies of learned societies amidst the evolving landscape of scholarly publishing. It also aims to alert other stakeholders, including researchers, librarians and funders, to the potential adverse consequences of this changing landscape for the health of learned societies and the academic disciplines they represent.

Literature review

Learned societies are charitable organizations which support academics and the dissemination of knowledge within a particular subject area or academic field. Publishing enables learned societies to further their missions by sharing knowledge,2 and, while only a minority publish peer-reviewed scholarly journals, societies are widely acknowledged as playing an important role in the scholarly publishing landscape.3

Scholarly publishing has been undertaken by societies for centuries, with the first scientific journals being established in the seventeenth century. Fyfe has charted the history of society publishing in the UK, including the entry and growing dominance of commercial publishers within the publishing ecosystem since the end of World War II.4 This period saw scholarly publishing shift from an often-subsidized mechanism for knowledge dissemination, to one which could generate a profit. In 2009 there were estimated to be at least 1,000 learned and professional societies in the UK5 but a 2015 study by Johnson and Fosci identified only 279 learned societies that published peer-reviewed academic content.6 However, for this subset of societies, publishing can be highly significant in revenue terms.7 While some societies maintain publishing programmes at a financial loss,8 Treadway and Greaves’ 2023 study of around 30 UK learned societies found that journal publishing represented between 30% and 70% of their income.9

The surpluses generated from society publishing are used to support other activities that serve their communities. These activities may include career development support for members, support programmes for students and early career researchers, educational resources and grants to support member activities, organizing scientific meetings, bringing evidence to decision makers to support evidence-based policy and running discipline-based special interest groups.10

Publishing is just one aspect of societies’ activities, with many affected by the loss of access to premises and revenue from events during the Covid-19 pandemic, in addition to moving to online and digital operations.11 In a publishing context, rising costs and technical requirements in a digital era have been accompanied by changes in how journals are sold, from individual copies and institutional subscriptions to vast multi-year collections and transformative agreements with library consortia. Negotiating such agreements poses a particular challenge for smaller society publishers with limited resources and market reach.12 Open access requirements through funder mandates and initiatives such as Plan S further complicate matters,13 while growth in preprints also poses a threat to subscription revenues.14 As a result, the publishing landscape has become increasingly diverse and complex, with Wise and Estelle identifying 27 different publishing models relevant to society publishers.15 Meanwhile, the scholarly journals market as a whole is shrinking in real terms,16 while open access publishing is increasingly dominated by large commercial publishers and has been characterized by Shu and Larivière as an ‘oligopolistic market’.17

One way for societies to ease these challenges is to outsource some or all publishing activity to a third-party publisher, traditionally one of the large commercial players or an established university press. The benefits of outsourcing include reduced technology and infrastructure costs, a predictable and stable income, access to deals which may otherwise be unavailable to a small society publisher, global reach, access to expertise, improved marketing and communication and support with transitioning to open access publishing. Yet outsourcing can bring its own challenges, including loss of editorial control, reputational risk, negative perceptions from others around independence and ownership of journals, limited influence and control over technology and software, and competition from other journals published by the partner.18

The literature suggests this is a uniquely challenging time for learned society publishers. On the one hand, recent calls for a ‘community-based scholarly communication system’ and growing concerns over research integrity may open up new opportunities for societies to maintain a leading role in publishing.19 On the other, as Fyfe has argued, ‘societies with a clear local or national identity may not be the best structures to manage a future of science communication that should be open, diverse and equitable and operate on a global scale’.20 By analysing the publishing models UK learned societies have adopted in recent years, and assessing the impact on their revenues, this article aims to help society leaders and other stakeholders make informed decisions about their future role in scholarly communication.

Methodology

For the present study, we began with a dataset of 279 UK-based learned societies previously identified in 2015 as having ‘an active peer-reviewed publication (i.e. academic journals or conference proceedings with an ISSN)’.21 Prepared as part of a Universities UK project to monitor the transition to open access in the United Kingdom, the dataset listed publicly available information on each society’s number of published journals, publishing partner(s) where publishing was outsourced and total incoming resources (based on their most recent published financial statements).22 Societies were also categorized by disciplinary focus using the classifications adopted by the UK’s Research Excellence Framework (REF), i.e. disciplines falling under panel A (which equates, roughly speaking, to medicine and life sciences), panel B (maths, physics, natural sciences, and engineering), panel C (social sciences), and panel D (arts and humanities).23 This classification is necessarily subjective and should be considered indicative only, as several societies have a multidisciplinary focus.

For the purposes of updating the dataset, we began by capturing the latest publicly available information for each of the original 279 learned societies. We first identified whether each learned society was still in existence, as evidenced by an active website and/or continuing registration with either the Charity Commission for England and Wales or the Office of the Scottish Charity Regulator. Where no such evidence could be found, or definitive evidence was found that a society was no longer in existence, the society was listed as ‘dissolved’. For those learned societies still in existence, we identified if they still had ‘an active peer-reviewed publication’ and, if so, recorded how many peer-reviewed publications were associated with the society, retaining the numbering system from the 2015 dataset (1, 2 or 3+). Where publishing of one or more peer-reviewed journals was outsourced, we captured the publisher(s) names. We used this data to assign each society to a publishing model in both 2015 and 2023, using the following typology:

• wholly self-published

• self-published and outsourced to one publisher

• self-published and outsourced to multiple publishers

• wholly outsourced to one publisher

• wholly outsourced to multiple publishers.

Updated financial information on total incoming resources and (where available) income from publishing was initially sought only for the 50 largest societies by revenue. As this sample was found to be dominated by societies in the life and natural sciences, additional financial data was sought for the ten largest societies in social sciences and the five largest in the arts and humanities to provide insights into a wider range of disciplines. In all cases this information was derived from statutory financial statements since all societies, whether registered charities or companies, are required to provide this information annually, and make it publicly available. Statutory financial statements must be prepared in accordance with generally accepted accounting practice (GAAP), ensuring some degree of comparability between societies.

For analysis purposes we introduced the concept of a ‘publishing relationship’ which identified each unique publishing relationship between a society and a publishing partner (in the format society-publishing partner), regardless of the number of journals published within that publishing relationship. For example, a society self-publishing two journals (society-society) was counted as one publishing relationship; a society self-publishing one journal (society-society) and outsourcing one journal to publisher A (society-publisher A) and two journals to publisher B (society-publisher B) was counted as three publishing relationships.

Findings

Of the original 279 UK-based learned societies identified in 2015, two were removed as having been included in error, three had been dissolved and seven had ceased publication. Figure 1 shows the numbers of societies by discipline, indicating that most societies with peer-reviewed publications are found in medicine and life sciences and arts and humanities. There are remarkably few changes in the numbers of journals published, with the exception of the medicine and life sciences. In these disciplines (which have typically seen the highest levels of open access adoption)26 a number of societies publishing a single journal in 2015 had launched additional titles by 2023. Review of these societies’ web pages confirms that the new titles were open access journals launched alongside a society’s existing subscription or hybrid journal.

Figure 1

A notable finding is the much higher number of societies which have ceased to publish peer-reviewed journals in the social sciences (seven) compared with the other three discipline groupings (three in total). This finding appears to bear out the cautionary observation made in 2015 by Johnson and Fosci, ‘The challenge of OA affects societies across all discipline areas, but perhaps most in jeopardy are small and medium-sized societies in the social sciences, which are often highly dependent on publishing revenues, but lack the scale and the external research funding to allow them to experiment or make the transition to a gold OA model’.27

The decline of self-publication

The Sankey diagram in Figure 2 illustrates changes in the publishing model used by the 277 societies in our dataset between 2015 and 2023. The most striking finding is a rapid decline in the number of wholly self-published societies, from 68 in 2015 to only 44 in 2023. The bulk of this decline relates to 17 formerly self-published societies who by 2023 had moved to an outsourced model with a single publishing partner. A further five of the self-published societies in 2015 have ceased publication entirely, one has been dissolved and two have chosen to combine self-publication and outsourcing. The pattern of societies ceasing to publish journals in their own right holds true across the disciplinary spectrum, with eight such societies in medicine and life sciences, seven in engineering and natural sciences, three in social sciences and five in arts and humanities.

Figure 2

A complex outsourcing landscape

Another trend observable in Figure 2 is the growing tendency for societies to outsource to multiple publishers rather than having a single partner. In 2015 only five societies outsourced their publishing to multiple partners and only two combined self-publishing with outsourcing. By 2023, these figures had roughly tripled in both cases, to 17 and six societies, respectively. In total, the number of outsourcing relationships identified in the dataset has increased from 214 in 2015 to 242 in 2023.

Table 1 shows how the position of key publishing partners for UK societies has changed over the period 2015 to 2023. Wiley has seen a reduction in its number of society partners while Taylor & Francis has seen a significant increase following its 2015 acquisition of Maney Publishing. Cambridge University Press (CUP), Oxford University Press (OUP), Sage and Elsevier have all increased the number of societies they work with. However, as societies are now more likely to outsource to multiple publishers rather than a single partner, we can infer that the financial value of each publisher-society relationship is likely to be lower in 2023 than it was in 2015. As a result, what looks like growth in large publishers’ market share in Table 1 may simply be a case of running to stand still.

Table 1

Publishers with 10+ relationships with UK learned societies – number of partnerships in 2015 and 2023

Publisher	Number of UK society partners (2015)	Number of UK society partners (2023)	Change
Wiley	59	53	–6
Taylor & Francis (including Maney Publishing in 2023)	37	53	16
Cambridge University Press	32	42	10
Oxford University Press	23	29	6
Maney Publishing (pre-acquisition by Taylor & Francis in 2015)	17	–	–17
Elsevier	14	17	3
SAGE	13	21	8

Figure 3 provides further context for this data, showing the publishing relationships for all 277 societies in 2015 and 2023. It indicates that, while the journals market as a whole is increasingly dominated by a small number of commercial players,28 the trend is towards greater diversity of publishing partners for UK learned societies.

Figure 3

The number of different partners used by UK learned societies rose from 17 in 2015 to 25 in 2023 and, as Figure 4 illustrates, this increase is being driven primarily by partnerships with university presses and, to a lesser extent, other societies, rather than new commercial entrants. In total, a dozen new publishers have entered the society market since 2015, while two commercial publishers (Bloomsbury and Emerald) have exited and two (Maney and Co-Action) have been acquired by Taylor & Francis, reflecting significant market consolidation over the period considered in our dataset.29 Among the new entrants, two (Frontiers and Wolters Kluwer) are commercial, three are university presses (UCL, Bristol and Chicago) and six are existing society publishers. Significantly, all 12 of these new entrants only partner with a single society at present.

Figure 4

Revenues lagging inflation

To aid understanding of the drivers of change within the society publishing market we also sought to assess how societies’ revenues from publishing have evolved since 2014. Only 21 of the 50 largest societies by revenue in our dataset disclose their publishing revenues, and these figures are subject to a number of significant caveats, as outlined in the methodology section. Nevertheless, the data presented in Figure 5 allows some broad conclusions to be drawn as to the health of UK society publishing. Firstly, most societies (15 out of 21) have suffered a real-terms decline in their revenues from publishing over the last decade or so. While some fluctuations in publishing revenues may be attributable to foreign exchange movements, the British pound weakened significantly against a basket of international currencies between 2014 and 2022.30 This weakening would have served to increase publishing revenues generated in other currencies on their translation into pound sterling, and so currency movements do not account for the decline and may, in fact, have masked its severity. In some cases, an apparent decline may also reflect a shift from self-publication (where a society’s financial statements disclose the gross revenues from publishing) to an outsourced relationship (where a society only reports the net revenues received from its partner). However, of the five societies reporting the steepest decline in their publication revenues only one had moved from self-publication to outsourcing, with one remaining self-published and the other three outsourcing to the same partners as in 2023. Notably, all five are in medicine and life sciences, being the disciplines in which open access publication is most widely adopted.31

By contrast, of the six societies which have succeeded in growing their revenues above inflation, four are in engineering and natural sciences (the Royal Society of Chemistry, the Institute of Physics, the Royal Society and the Institution of Engineering and Technology). The other two are the British Academy, which derives only 1% of its income from publishing, and the Society of Endocrinology, for which the most recent publicly available financial data relates to the year ended 31 December 2020, meaning it may be distorted by the Covid-19 pandemic. Figure 5 illustrates a correlation between a society’s reliance on its publishing revenue and its ability to grow those revenues, though the small sample size and wide variation in the nature of the societies considered means such findings must be treated with caution. A review of the underlying data further reinforces the importance of scale: the ten largest societies disclosing publishing revenues saw a median growth of 17% (mean growth: 8%), while the 11 smaller societies saw a median decline of 20% (mean decline: 22%).

Figure 5

We also reviewed changes in publication revenues for the ten largest social science societies, and the five largest in arts and humanities. Trends within these disciplines are harder to discern given the small sample size, inconsistent disclosure of publication revenues and greater significance of monograph publication for some of these societies. However, as noted by Johnson and Fosci, social science societies appear particularly exposed to changes in the marketplace as they are more likely to be dependent on a single journal for a significant share of their revenues.32

Discussion

The overall picture emerging from our analysis is a sobering one, albeit with some bright spots. As a whole, the significance of UK learned society publishers within global scholarly communication is waning as the centre of gravity of the world’s research shifts eastwards and market conditions favour publishing models focussed on quantity rather than quality.33 The very largest UK society publishers in the natural and life sciences have been able to achieve sustained growth and journal reputation and prestige remain the most important determinants of authors’ behaviour.34 However, those societies lacking scale or brand recognition have seen their revenues progressively squeezed.

Amongst all this, the influence of the global pandemic on the publishing landscape cannot be ignored. On the one hand, in many disciplines restrictions resulted in reductions or cessation of research due to lack of access to premises, people and materials. On the other, the increased time available to many researchers and the urgency of the research questions posed by SARS-CoV-2 led to a sharp increase in published output, much of it openly accessible.35 The trend towards increased scale and complexity in scholarly publishing long predates the pandemic but it undoubtedly led to its acceleration, while simultaneously weakening the financial position of the academic institutions and libraries on which publishers depend.36

Not apparent in our data, but amply evidenced in the literature, are the additional costs associated with the transition to open access, which are likely to have put pressure on societies’ margins from publishing even in cases where revenue has grown.37 While many societies cite these costs as necessitating a transition from self-publication to outsourcing, our findings suggest such a move may take them only as far as the fire from the frying pan. Of eight societies that were self-published in both 2015 and 2023 and for which we gathered data on publication revenues, all but one were able to grow their revenues from publishing, with a median growth figure of 21%, roughly in line with inflation (mean growth: 12%). By contrast, of the seven societies that had outsourced their publishing in both 2015 and 2023, five saw an outright decline in their revenues, with a median reduction across all seven of 30% (mean decline: 31%).

The numbers considered in this analysis are small and there are a range of factors which may explain the difference in observed financial performance between self-published and outsourced societies. Large societies are more likely to self-publish, for example, so the difference may reflect the benefits of scale rather than self-publication. The relative decline in revenues for societies with publishing partners may partly reflect the generous terms some societies were able to negotiate with commercial publishers in the early 2010s, which have not been sustained into the 2020s. Furthermore, many societies, as not-for-profit entities committed to the dissemination of knowledge, may consider a reduction in revenues an acceptable price to pay for increased access to the research published in their journals. Nevertheless, given the difficulty of extricating a society from a big deal and the falling value attached to society journals, these data should give independent society publishers contemplating a move towards outsourcing pause for thought.

Meanwhile, the appearance of so many new publishing partners for societies would appear to be a double-edged sword. The willingness of new university presses and other societies to act as publishing partners is to be welcomed, but their inability or unwillingness (to date) to take on multiple society partners is likely to be attributable to the same reason that commercial players have not made further inroads into the UK society market. A society journal is no longer the ticket to riches that it once was, and the fear on all sides is that as the transition to open access proceeds its value will be only further eroded. A key question over the coming years will be which of the many ‘single society’ publishing partners quietly exit the UK society market and which seek to increase their presence.

Conclusion

One conclusion from our work is that change in the society publishing landscape happens slowly. In 2015, 175 societies, or 63% of the total, published a single peer-reviewed journal. By 2023, 157 of these 175 societies continued to publish a single journal, despite all the changes in the external environment over this period. A second conclusion is that, though slow, change is occurring. This is evident in the declining number of self-publishing societies, the increased complexity of the outsourcing landscape and the failure of most societies’ publishing revenues to keep pace with inflation.

Within this context, inertia poses the greatest threat to many societies, with a progressive erosion of their subscription revenues the most likely result. The role played by open access in these changes cannot be discounted. It would appear no coincidence that societies in medicine and life sciences, where open access publication is most prevalent, are also those that have seen the sharpest decline in revenues. Societies in engineering, the natural sciences and the social sciences look set to tread a similar path over the coming years. Societies in the humanities – mostly small, with low levels of reliance on publishing revenues and the slowest rate of progress towards open access – are perhaps the least exposed to changes in the publishing landscape, but low-output, highly selective journals in all disciplines face an uncertain future.

For policymakers and librarians, this article sheds light on an unintended consequence of the transition to open access. In the context of growing concerns over research integrity, now further exacerbated by the rise of generative artificial intelligence, the role learned societies can play as arbiters of quality in their disciplines appears increasingly important. Yet societies’ ability to play this role is increasingly threatened by the erosion of their publishing revenues and a narrow focus on driving down the cost per publication. In the future, more nuanced approaches to assessing the ‘value’ provided by publishers may be needed, considering a broader range of factors including a publisher’s community governance, openness, equity, (not-)for-profit status, and potentially indicators of publication integrity such as rejection rates.

Finally, our findings should provide encouragement to society publishers that outsourcing to a large commercial publisher is not the only viable strategy and may even lead to worse financial outcomes than remaining independent. Societies are now able to access a much wider range of potential publishing partners, collaborators and service providers than they were a decade ago, including other societies and university presses whose missions and values are more likely to align with their own. Yet leveraging these opportunities is likely to require an investment of management time and financial resources that will not be feasible or even desirable for all societies. Where publishing is a marginal part of a society’s activities or its appetite for risk is low, the logical approach may still be to divest what is an increasingly complex operation offering uncertain future returns. But for those societies with the means and the will to publish journals in their own right, this study bolsters the case for retaining, or even reclaiming, their independence.

Data accessibility statement

This article references the following publicly available datasets:

R. Johnson et al., “Dataset of UK learned society publishers 2015–2023,” Dataset, Zenodo last modified May 12, 2024, Version v3, https://zenodo.org/doi/10.5281/zenodo.10257806.

Rob Johnson, “Financial Sustainability of UK Learned Societies – List of UK Learned Societies,” Dataset, figshare, 2015, https://doi.org/10.6084/m9.figshare.1476266.v1.

Abbreviations and Acronyms

A list of the abbreviations and acronyms used in this and other Insights articles can be accessed here – click on the URL and then select the ‘full list of industry A&As’ link: http://www.uksg.org/publications#aa.

Competing interests

The authors are consultants working in the field of research and scholarly communication and have provided consultancy services to a number of the publishers and societies mentioned in this article.