What benefits do traditional publishers provide?

Jonathan Aldrich
ACM SIGCHI
Published in
9 min readDec 6, 2023

We invited Jonathan Aldrich, SIG Governing Board Publications Advisor, to continue the conversation we/SIGCHI started with our first post on ACM OPEN, as we aim to ensure our SIG’s awareness and understanding of the implications of this upcoming transition to 100% Open Access. Thank you, Jonathan!

The Open Access movement is transforming academic publishing. Based on grassroots petitions and pressure from funders, publishers are moving to a Gold Open Access model, in which every paper is freely available to any reader, immediately after being published. Open Access will transform science by ensuring that knowledge is available to anyone who wants it, regardless of ability to pay.

Publishing papers does have a cost, however. Open Access requires those costs to be recouped up front, when a paper is published, rather than paid for by readers over time. The advent of Open Access has made the cost of publishing more salient to authors, because one solution to paying for publishing is to charge authors Article Processing Charges, or APCs for short. APCs vary enormously: publishing a Nature article through Springer, a for-profit publisher, will cost the authors $11,690; publishing a conference paper as a member of the nonprofit Association for Computing Machinery (ACM) will cost authors $700; and publishing a paper through LIPIcs typically costs the author nothing, though the conference pays LIPIcs $60 per paper.

Why do these costs differ so much?

What traditional publishers provide

Traditional publishers are organizations that curate journals and sometimes conferences, coordinate reviewing for them, typeset and edit articles (sometimes left to authors now), and distribute articles (historically by printing them, now usually via digital libraries). They provide a number of benefits to the scientific community. We’ll look at these benefits with a particular eye to those provided by the ACM, the leading professional organization for computer scientists:

(1) Stability. You want an organization you can count on to make your papers available in 100 years. Publishers have contractual backups: if the ACM goes under, another organization will step in with a commitment to make the papers available forever.

(2) Maintenance. Formats don’t change quickly, but remember the online format used to be .ps (or .ps.Z). Today it is PDF, and tomorrow it will be HTML. Even paper content changes, e.g. when people change their names both paper authors and references have to be updated. A publisher should maintain its papers in the latest formats and make updates to their content and metadata as necessary.

(3) Digital Library Platform. It’s not just about serving PDFs — publishers provide extensive metadata and analytics about their hosted papers. This supports searching for papers in a digital library, linking to cited papers, finding papers that cite a paper, and assessing the impact of a paper or author. One might be tempted to rely on free commercial services such as Google Scholar for such purposes, but although these services can be convenient, it is unwise for the community to rely on them. Not only can commercial search results be tainted by commercial interests, such services can go away in the blink of an eye — as in fact the former Google Code service and countless others have in the past.

(4) Quality. Publishers like the ACM enforce high standards for publication and review. In contrast, the level of quality assurance done at services such as arXiv (which may seem “free” but require subsidies from libraries, the goverment, and others) is much lower, and it shows.

(5) Accountability. What if someone plagiarizes your paper? What if a published paper plagiarizes others? What if a paper has falsified data? ACM will investigate and take what action they can.

(6) Promotion. ACM works to make sure its journals are indexed by Clarivate. This is important to a lot of people for whom only indexed journals “count” towards promotion, grants, etc. The bean counting is unfortunate but it’s a reality that many of our colleagues deal with.

(7) Services. ACM is working on various services to help with reviewing, including conflict detection, finding expert reviewers, spreading reviewing workload more easily. Publishers are well-positioned to help with this.

Comparison: ACM vs. LIPIcs

How do the services that ACM provides compare to those provided by alternative, cheaper publishers, such as LIPIcs? We should begin with a caveat that LIPIcs is funded by the German government and state governments, so its financial needs and costs are not fully comparable to those of ACM. With that caveat, however, let’s use the above categories to analyze the differences:

(0) Cost. LIPIcs is much cheaper, $60 vs. $700 for conference papers (the ACM member price)

(1) Stability. I’d give the edge to ACM as it has been around longer, but both pay backup orgs for long term preservation, so it’s close.

(2) Maintenance. Clear ACM win. It has an active program for making HTML format available and an author name change policy. I couldn’t find anything for LIPIcs in either category.

(3) Digital Library Platform. LIPIcs has full-text search and very limited metadata per paper. The ACM DL has far better search facilities, cross-citation links, and data to assess a paper’s impact. Clear ACM win.

(4) Quality. Both have rigorous processes for selecting conferences. I’d say ACM has an edge from its longstanding reputation, but let’s call this at most a small edge, as there are great conferences published by LIPIcs.

(5) Accountability. ACM win. It provides dedicated staff and clear processes for reporting and investigating publications violations. I could find nothing on this at the LIPIcs site.

(6) Promotion. ACM win — it has been actively promoting its journals with Clarivate, resulting in (e.g.) PACMPL being indexed this year. LIPIcs mentions the possibility of acceptance by Clarivate as an acceptance criterion, but they don’t say they are going to advocate for your journal like ACM does for its journals — it’s probably going to be on the conference. And this is hard.

(7) Services. ACM win — lots of initiatives in this area, and there is no indication that LIPIcs is doing similar things.

So in summary, the community does ultimately pay more for ACM as a publisher, but it also gets more. I don’t have anything against LIPIcs, to be clear — I’m the PC chair of ECOOP 2024, which is published by LIPIcs. LIPIcs is a fine match for ECOOP, but the tradeoffs are quite different from using ACM as a publisher.

Who actually pays Article Processing Charges (APCs)? Mostly, not you!

While the above comparisons may be instructive, it’s important to keep a perspective. The APC is the “sticker price” of publishing a paper in the Open Access era. But in the payment models most likely to dominate, individual researchers will not pay APCs.

The ACM, along with other publishers, is offering “Read-and-Publish” agreements to universities, companies, and governments that allow members of the organization to publish open access articles for free. The program is called ACM OPEN, and pricing is based on the number of papers published by the organization each year. ACM OPEN has been a thing since 2020, and already 35% of ACM articles are published under ACM OPEN agreements. The number of subscribing organizations continues to rise. It is ACM’s expressed goal to “flip” its entire library to be open access as of January 1, 2026, a little more than 2 years from now. At that point, all papers (past and present) will be freely available to the public, and all papers will have to be published via ACM OPEN, an APC, or an APC waiver. It’s becoming clear that most people will not have to pay an APC:

  • High-resource institutions such as top research universities have been among the first to subscribe to ACM OPEN. They can see that it is a good deal for them because open access articles are downloaded 3x as much and cited 66% more. That means more impact for the work of researchers at those institutions — and it’s worth paying for. When the institution pays for a library subscription, authors don’t pay APCs.
  • Low-resource institutions typically don’t publish many papers. For an institution in the lowest “publishing tier,” i.e. that publishes a handful of papers each year, ACM OPEN subscriptions will be cheaper than their current costs to Subscribe-to-Read. Thus it is both affordable and worthwhile for these institutions to subscribe also.
  • APCs are waived for authors based in low-income countries. For authors in lower-middle income countries, there is a 50% discount on APCs.
  • Even aside from waivers and discounts on APCs, the ACM is offering good deals to institutions in economically developing countries to bring them on board. The details are negotiated and depend on individual circumstances, but as with low-resource institutions, ACM OPEN may be available to universities in some economically developing countries at the same price they currently pay to read papers. These deals are very likely to be accepted by the “flip” in 2026, since they increase benefits without increasing costs.
  • There are some countries, such as China, where many institutions have not yet signed on to ACM OPEN (though a few top Chinese universities have done so). However, in China, funding agencies are willing to provide grants with money for APCs included, as this increases the reach of articles published by Chinese researchers. Even in middle income countries such as China, the APC for a conference paper is a small fraction of the cost of supporting a graduate student. So while the cost is not insignificant, it is not a huge burden to research, either, and many researchers would happily trade the increased visibility from open access publishing for this additional cost.
  • If none of the above applies, authors can request APC waivers based on their individual circumstances. Although policy is being finalized, we expect that waivers will be routinely granted to authors who lack external funding, are not highly-paid consultants, and who are from low resource universities or have no institutional affiliation.

Based on the above, we do not expect APCs will be a substantial barrier to publishing with the ACM, even for authors with few resources.

Could ACM’s APCs be lower?

ACM is a nonprofit — its goal is not to make money. Therefore, both APCs and ACM OPEN fees to institutions are designed to be revenue neutral: after the “flip” to open access (and in fact, during the process as well) ACM will get no more money per paper than it does today.

ACM’s APCs match the costs of publishing pretty closely: the cost per conference paper was reported as $709, almost exactly ACM’s member APC, in the 2019 overview of ACM Publishing Finances (note that this figure is 4 years old by now, so realistically the costs below have increased some due to inflation). ACM OPEN prices are calculated so that the organization would break even if all organizations subscribed rather than paying APCs.

Using the figures in the overview article cited above, how does that $709 cost from 2019 break down? By my calculations, it works out approximately as follows.

Publishing cost: $410

  • Reviewing software (e.g. HotCRP): $44
  • Production services (compile and produce proceedings): $49
  • Plagiarism detection, DOIs, etc: $3
  • Core staff: $22
  • Overhead & support staff (legal, HR, marketing): $34
  • Return to SIGs (for their conference publishing expenses): $258

Digital library cost: $299

  • Core software platform (maintenance, improvements): $77
  • Content delivery (servers, etc.): $52
  • Interactive services (comments, notifications, email): $13
  • Statistics, analytics, and other value add: $41
  • Content preservation (if ACM goes away, what happens?): $2
  • Core staff: $54
  • Overhead & support staff: $60

The “return to SIGs” category deserves a bit more explanation. Historically it came about as a way to return some of the income from conference proceedings subscriptions to the SIGs, since the SIGs run the conference reviewing process that produces proceedings articles. The money covers expenses associated with reviewing and publishing papers that are associated with the conferences the SIGs run, thereby lowering conference registration fees. Any surplus is used by the SIGs to benefit their community, such as by providing funds to help students attend conferences.

When you look at the budget breakdown, it’s not obvious what can easily be cut. Certainly the ACM could decline to return money to the SIGs for their conference publishing expenses; then the member APC could perhaps be $450 instead of $700. But, then the SIGs would have to raise conference registration fees to cover the costs of publishing to them and to support the other things they do. Other expenses could be trimmed, but there would be corresponding effects on the quality of ACM’s various services, as described above.

Ultimately, it’s up to you

Hopefully this post has helped flesh out what ACM does as a publisher and the value provided by its publishing services. I think the ACM provides quite a lot of value to the CS community! Ultimately, however, ACM is a member-led organization. What publishing services do you think ACM should provide going forward?

Many, many books
Photo by 愚木混株 cdd20 on Unsplash

--

--

Responses (1)