ACM, SIGCHI, and the Economics of Open Access Publishing

tldr;

• Publishing has material costs, and understanding the economics of non-profit publishing is important for meaningful discussions about open access
• Scholarly societies are non-profit organisations that sustain important community activities by reinvesting revenue from various sources, including publishing
• If we want universal gold open access, we need to explore other economic models or reduce community activities

Estimated Reading Time: 10 Minutes

---

ACM could transition to universal gold open access and make all content in the ACM Digital Library freely available.  This would have significant implications for the ACM as a scholarly society. For example, whether we are able to continue financially supporting community activities like student travel grants, career development events, underwriting conferences,  and other good works. This post explores the finances of non-profit publishing, addresses common misconceptions about publishing with scholarly societies, and what this means for our values as a community.

Publishing has real costs in terms of software, storage, bandwidth, and professional staff. To fund universal open access, organisations like ACM would need to replace publication revenue streams or reduce community activities. One approach would be to increase conference fees to cover these costs. As an example, SOUPS increased registration fees by 86% from 2013 to 2019 after transitioning to open access with USENIX in 2014^[1].  Alternatively, community initiatives that are not economically self-sustaining like student travel grants, regional development activities, educational events, and career development could be cut.  For example, IEEE spent $15 million on public engagement and outreach in 2018 as part of their “public imperatives” program^[2]. These are difficult tradeoffs that require careful consideration.

A common misconception about revenue in scholarly society publishing is that “someone” must be profiting from the members’ hard work.  Before this post takes a deep-dive into the finances of open access and non-profit organisations, it’s important to recognise that “profit” it not part of the equation. All of the organisations that I’ll be discussing (ACM, arXiv, IEEE, and USENIX) generate an annual revenue that covers operating costs and a surplus that is reinvested in the community and/or held as assets.  These assets are crucial to secure on-going financial commitments and underwrite events. Building up these assets is necessary to underwrite increased financial liabilities as these organisations grow^[3].  

This financial structure makes it possible for scholarly societies to support their communities and insulate members from risk.  For example, scholarly societies take on financial and legal liability for conferences and events by securing them against society holdings.  This provision of financial security is a critical service. Many of the events and activities we enjoy as members would not be possible without the bankroll of a well-funded society.  This makes it possible to plan on longer timescales, scale up events, and take financial risks such as organising events in new regions.

What do publishers provide?

Publishing scholarly work is more complex than simply hosting PDFs^[4].  While we could all choose to host un-reviewed content on our personal websites^[5], there is value added by publishing with a professional organisation.  A publisher should provide three things: discoverability, availability, and persistence.  Providing these costs money in terms of staff, volunteer support, infrastructure, systems, servers, and bandwidth.  

Discoverability – Content must be discoverable

Making content discoverable requires not only software and hosting, but also paid services like DOIs^[6], and strategic planning and enforcement of standards around metadata, library structure, visual identity, and classification. Organisations like ACM and IEEE achieve this through professional staff who prepare author submitted content for their digital libraries, assign DOIs, verify content and metadata, and manage document release operations.  arXiv follows a different model, where this work is distributed to authors and volunteer subject moderators, and some paid services (like DOIs) are excluded and arXiv specific identifiers are used instead.

Availability – Content must be available for download

Serving articles has a material expense in terms of bandwidth and storage. How articles are made “available” varies significantly.  Organisations like ACM and IEEE put the majority of their content behind a paywall by default, although both have a range of options for open access publication.  arXiv makes all content freely available by default, but requests annual membership fees based on usage^[7]. USENIX follows a different model, where conference budgets must generate enough revenue to cover the cost of gold open access publishing.

Persistence – Content must be preserved

It is also important that content is published in resilient formats that will endure beyond the lifetime of the publishing organisation.  ACM has recently transitioned to an XML based JATS/BITS archival format and from 2020 onwards publications are no longer coupled to reading formats like PDFs.  This future-proofs publications, ensuring content can be rendered in PDFs, HTML, ePub, and formats that have not yet been imagined.  This initiative required significant coordinated investment in staff,  volunteer support, software, and infrastructure.  There are also long-term fail-safe preservation costs to ensure the availability of content in the event that ACM’s platforms fail or ACM ceases to exist through services like Portico.

How does ACM compare to other organisations?

The following table shows relevant operating costs and revenues for ACM, arXiv, IEEE, and USENIX.  This analysis is based on publicly available annual reporting for Financial Year 2018 for ACM and IEEE, Calendar Year 2018 for arXiv, and Financial Year 2017 for USENIX.  All values are given in thousands of US Dollars.  Citations for relevant documents and additional comments are given as footnotes.  

Table Format

Revenue         (% total revenue)

Expenses         

Profit/loss         (% total revenue)

	ACM[8]	arXiv[9]	IEEE[10]	USENIX[11]
Overall[12]	$80,827K -$73,427K +$7,400K (9%)	$1,411K -$1,338K +$73K[13] (5%)	$535,711K -$503,652K +$32,059K (6%)	$6,330K -$5,358K +$972K (15%)[14]
Publications[15]	$22,732K (28%) -$10,854K +$11,878K (15%)	— -$1,338K[16] -$1,338K (95%)	$225,514K (42%) -$193,382K +32,132 (6%)	—[17] — —
Activities	$36,526K (45%) -$33,697K +$2,829K (4%)	— — —	$245,957K (46%) -$196,821K +$49,136 (9%)	$5,064K (80%) -$4,554K +510K (8%)
Membership	$7,101K (9%) -$981K +$6,120K (8%)	$542K (38%) — +$542K (38%)	$63,999K[18] (12%) -$103,458K -$39,459K (7%)	$229K (4%) -$308K -$79K (1%)
Grants	$8,139K (10%) — +$8,139K (10%)	$796K (56%) — +$796K (56%)	— — —	— — —
Other[19]	$6,329K (8%) -$27,895K -$21,566K (27%)	$73K (5%) — +$73K (5%)	$241K (<1%) -$9,990K -9,526K (2%)	$1,037K (16%) -$496K +$541K (8%)
Assets	$123,289K[20]	$1,229K	$658,330K	$6,381K

 

Notes on this Table

This table gives an overview of organisations with different structures, in particular at different scales.  The work of scholarly publishers is similar to the work of software engineers in that scaling up can have an inverse economy of scale as complexity increases and coordination becomes more complicated.  The same is true of large events and conferences, where costs-per-head typically increase as an event scales up.

Each organisation captures volunteer effort and volunteer support differently.  In budget lines with significant loss, this usually represents an area of spendincg and investment in community activities. Because these organisations are all structured differently, how publishing costs are captured and what is included in “publishing operating costs” and “general operating costs” varies.  

ACM

• Organisation funded through publications, grants, membership fees, and conferences
• ACM Digital Library access through subscription (library funding) or personal access (membership, pay per article) with flexible green open access and paid gold open access
• Extensive community activity through ACM and Special Interest Groups (SIGS)

ACM uses a range of revenue streams to support their activities, which are implemented through ACM and Special Interest Groups (SIGs).  SIGs receive funding from ACM publications revenue based on downloads of sponsored publications and have significant flexibility in how they operate.  In ACM financials, reinvestment of revenue in the community is included in “Other” as programme support and through SIG budgets (not reported separately).

arXiv

• Organisation funded through grants, donations, and institutional membership (library funding)
• All publications available gold open access
• No activity outside publishing

arXiv is supported primarily through grant income and institutional membership with fees based on usage.  arXiv is a lean organisation focused solely on sustaining its open access digital library.  ArXiv’s high level of activity and impact is a strong case for universal open access.

IEEE

• Organisation funded through publications, conferences, standards, and membership fees
• IEEE Xplore Digital Library access through subscription (library funding) or personal access (membership, pay per article) with green open access and paid gold open access
• Extensive community activity through IEEE and Societies

IEEE is a large scale organisation with a wide range of activities including publishing, conferences, standard development, career development, education, and policy.  IEEE is organised into Societies which also support these activities.  In IEEE financials, reinvestment of revenue in the community is captured under “Memberships” and a detailed breakdown of this spending can be seen on page 61 of the IEEE 2018 Annual Report.

USENIX

• Organisation funded through conference activity and membership fees
• All publications available gold open access
• Limited activity outside of conferences and publishing

USENIX is dedicated to open access publishing and all events sponsored by USENIX must cover the costs of open access publishing as part of conference budgets.  Conference budgets also include reinvestments in the community in line items such as travel grants. There are also “good works” captured under “Memberships.”

What Models Could We Explore Moving Forward?

There are a number of ways forward where an organisation like ACM could sustain open access publishing and it is likely that a hybrid of models will be needed.  It is also likely that change will need to be on the scale of years, to ensure that societies don’t become insolvent before transition to open access is complete.

Library Funded (Maintain Activity)

One approach to supporting open access while maintaining scholarly society activities is changing how subscription models work.  For example, Read and Publish (RAP) Agreements are a new subscription model that includes gold open access publishing alongside access to paywalled content.  RAP models allow for comparable revenue from library budgets and build in mitigation for lost revenue as an increasing proportion of a digital library becomes open.

Event Funded (Maintain Activity)

Another approach is to increase revenue from conferences, events, and standards.  For example, ACM currently depends on the $23M annual revenue from publications. As an illustration, ACM conferences and events make a profit of $3 million on a revenue of $37 million.  Assuming fixed costs, this revenue would need to increase by 62% to $60 million to fully replace publication revenue.

For example, increasing revenue by 62% for a large conference like CHI2018^[21] would mean increasing the $800 registration fee to $1400^[22] to achieve an overall revenue increase of 62%.  This is comparable to registration fees increases at conferences like SOUPS after transitioning to open access with USENIX.

Those with experience running a conference and managing conference budgets will know it can be difficult to break even, let alone substantially increase profits.  As a community, we may not be comfortable with increased fees and how this might affect the accessibility of our events.

Reduced Activity

Another approach to supporting open access would be to strip back society activities with a more moderate increase in conference revenue.  For example, replacing ACM’s $23M publication revenue could be done by increasing conference revenue by 30% to $60 million while cutting $12 million (15% of annual operating expenses) from spending on community projects.  Since much of ACM’s community activity is implemented through SIGs, this would likely have a significant impact on the kinds of activities SIGs could organise and support.  This is the “USENIX” model where open access publication is supported and there is limited activity outside of conference events.

Publish Only

Scholarly societies could instead maintain conference income at similar levels and radically strip back all other activities. This would also mean dramatic changes to a society’s remit, goals, and objectives.  This is the “arXiv model” of a lean organisation supported by volunteer effort, where there is no significant activity outside of publishing/digital archive.  In the case of ACM, this would likely mean dissolving all SIGs, dissolving boards and regional councils, discontinuing all educational activities, and cancelling all events that are not economically self-sustaining. Even a radically stripped back organisation requires revenue to function, and in the case of arXiv this is primarily through grants, donations, and institutional membership fees.    

Philanthropy

If an organisation could attract significant philanthropic donations it could support open access based on return on investment.  In 2018, ACM had an operating budget of $81 million.  Based on a conservative 4% return on assets annually, ACM would need an additional $2 billion in investments to cover existing operating costs, or an additional $575 million to replace publication revenue.

Values

It is important to discuss how we want our society to function and how we express and prioritise our values.  I hope that this blog post helps clarify the financial and organisational issues that must be considered when talking about open access in the context of scholarly societies.  For societies like ACM and IEEE, this will mean balancing some competing needs and priorities.  To conclude, I want to highlight a few of the key issues.

• As a community, are we willing to strip back our community activities and good works in favour of open access publication? While open access is crucially important, how do we balance this with other important issues that are currently financially sustained by scholarly society structure?
• As a community, how do we feel about subsidizing good works and community activities from revenue streams like publications and events? Given “who” is paying in these different models, are we comfortable with where the revenue is generated?
• Are we willing to give up the security and insulation from financial and legal risk that scholarly societies currently provide in favour of leaner organisations?
• How much value do we place on current models of publication that include peer review, standard visual identity, and metrics/indexing as part of how we measure quality?  These activities add overheads and costs to publishing, and as we change publishing revenues, we may need to reevaluate these issues. Career progression internationally is also tied up in these structures, and rejecting them entirely may not be feasible or desirable.

If you walk away with nothing else after reading this blog, I hope you have a deeper understanding of the complexity of non-profit publishing, the range of different approaches taken by different non-profit organisations, and the challenging decisions and trade-offs that we will have to make in the future if we change the organisational and economic models of scholarly societies.

---

Julie R. Williamson is an academic at the University of Glasgow and the current Vice President for publications for ACM SIGCHI.  She is actively involved in publication policy through work on the ACM Publications Board and the ACM Europe Council.

---