Author’s Note: I was invited to be a speaker for the Open Access Webinar Series by OhioLINK. The post below is a shorter version of the webinar lecture. I would like to thank Gwen Evans for the invitation and the webinar participants for listening.
I have been thinking a lot about downward pressure of pricing for scholarly publications.
The costs associated with publishing content in a professional way is part and parcel to the discussions on which business models should be associated with scholarly communications. Leaving behind the silly arguments about what is an acceptable profit margin or surplus, and who decides what kinds of activities one is allowed to do with those surpluses, I want to explore a bit about what I see as “publisher added value,” look at some numbers that are complicating the issues, and imagine a future world in which these services or activities cease to exist as we collectively race to the bottom in order to cut expenses.
Downward pricing pressure is happening regardless of the business model of the journal. Yes, the Plan S intention to cap Article Processing Charges (APCs) is currently fueling some cost speculation; but we also see pressure to keep year-on-year subscription price increases low, despite the fact that more services are required of publishers and there is more content to review and perhaps produce.
Open access (OA) journals have not been immune to criticisms on pricing strategies (here and here, for example). The main goal of the OA movement was to make more content accessible to all with liberal reuse rights. This is a very good talking point and one that — if better organized — would have resulted in a faster and more widespread adoption of OA than has happened. The lack of consensus or clear-cut answers on how widespread OA would work, be financially supported, and be sustainable in the long term, meant that these issues were largely ignored.
It should not have been any surprise that the largest commercial publishers would be able to roll out OA options for those who wanted it with ease — at a price. There is an increasingly better understanding that producing and disseminating scholarship does, in fact, cost money. Initial concerns that a universal flip to OA would only shift money around as opposed to reducing the amount of money put into the system, have been proven true.
The term “publisher added value” has become a pejorative way of dismissing publisher concerns about this downward pressure on pricing. That said, I do think it’s perfectly reasonable to ask, what do publishers do to scholarly content that adds value?
While I will not be giving you an itemized lists of the costs of publishing a journal article, I want to focus on some of the hidden costs associated with journals, regardless of whether they are OA or subscription journals. I also want to imagine what might happen if these tasks were no longer performed.
The peer view process has been in the crosshairs for the last few years. Authors complain of bias and the length of the process. Reviewers complain that the quality of papers is substandard and authors are lazy about making changes. Readers complain that flawed papers are still passing review. And editors complain that there aren’t enough people willing to perform quality peer review to cover the vast increases in the number of papers being submitted.
When you don’t get print journals that make a thud when they hit your desk, it can be easy to miss that the journals themselves are getting bigger. Further, journal papers are getting longer as publishers have relaxed page limits online. But there is a nasty little secret about that. We still pay for production services like copyediting, tagging, and typesetting by the page — even if we aren’t printing the journals.
I really can’t underestimate the impact that a large increase in submission rates can have on a journal:
- More editors are needed
- Larger reviewer pools are needed
- More staff are needed
- All submissions cost money, even the ones that aren’t accepted
For all journals that reject more papers than they accept, the cost of rejections is high. This is true for OA journals as well. A competitive APC must also subsidize the costs associated with rejected papers and papers published with waivers. That expense will depend on the acceptance rate of the journal.
In order to address concerns of quality and in response to highly public wrong doings in scholarship, new steps are taken such as Similarity Checks, the EQUATOR checklists, conflict of interest reviews, checks for excessive self-citations, checks for citations to retracted articles, and more. Journals are now starting to require data availability and compliance with FAIR data principles, all good for science and reproducibility, all requiring staff resources.
Even services such as ORCiD, built as a way to disambiguate individuals and their work, is being implemented in the submissions systems as one more check on whether a reviewer is who they say they are — an important step for publication managers to ensure the integrity of the peer review process and a direct response to the unthinkable — authors posing as reviewers and scamming the system.
Journals are also using forensic software to review graphics and images to ensure that they were not fudged.
These checks and balances are extremely time intensive and carry a considerable cost. They are also not perfect. As fraud detection tools have increased, the fraud perpetrators have become savvier at fooling the system. Every time a red flag goes up on these quality checks, an investigation of some sort ensues. It typically starts with a query to authors, again requiring time and expertise of journal staff and editors to properly communicate the findings and the resolution needed. Days, sometimes weeks, are needed with a series of back and forth emails to determine whether the author made an honest mistake that can be fixed or if the paper should be rejected. Of course, the fun doesn’t stop at rejection. Often further investigations, possibly involving other journals and an author’s institution ensue.
The expectation is that journals will identify fraud and careless mistakes as part of the peer review process. There are assumptions that this is all quick and inexpensive, when in fact, it requires specialized training of staff and tools or memberships that carry expense. Ethics training of staff is an ongoing commitment as new ways of perpetrating fraud are constantly being discovered. Each high-profile retraction leads to process changes and additional reviews by journal staff and editors everywhere.
It’s not perfect, but it’s nowhere close to easy and it will only continue to get more complicated. Professionals whose job it is to ensure that fraud is discovered and that public safety is protected get a little bent out of shape when there are calls to abolish peer review as we know it. I have had to talk many an editorial coordinator off the ledge when a blog post, or worse a mainstream news article, questions whether what they do for 40-plus hours a week adds value to scholarship.
Pressure to decrease costs may mean that resources will no longer be spent on these activities. Worse, a business decision could be made that because the vast majority of authors and readers (1) don’t know we do these things and/or (2) don’t see a value in them, they should cease. I believe that these activities are of value to the overall scholarly community, but if paying for it is an issue, then perhaps the research institutions would need to insert a manuscript review process for papers prior to posting or journal submission.
Beyond the nuts and bolts of quality control, the facilitation of peer review is an onerous task. There have been calls for posting papers on preprint servers and expecting the community to spontaneously conduct peer review. I have my doubts that this will ever work. Publicly commenting on scholarly content has never been popular.
The facilitation of peer review — with a mechanism for editors to invite reviewers, follow up with reviewers, and successfully acquire reviews — does not happen by accident and removing the facilitators, in most cases journals staff, means that it won’t happen.
When discussing the cost of publishing a journal, what we hear most is that the peer review costs nothing because the reviewers aren’t being paid. Clearly publishers have not done a good enough job communicating what happens behind the scenes. Unless they are journal editors, the average author has no desire to look behind the curtain, which is understandable given the small percentage of their time spent being an author. However, before calls for defunding or eliminating formal peer review altogether come to fruition, the ramifications of doing so must be fully explored.
Moving on from peer review, the production processes for journals are also often dismissed as unnecessary or woefully undervalued; however, dissemination and discovery of content is highly dependent on production work. Standardizing content and tagging for maximum interoperability is important for taking advantage of the rich offerings in content discovery.
Crossref reference linking will not work well if references aren’t formatted and missing information provided. Unstructured, incomplete, and poorly tagged references are minimally useful to the community. There are two options for fixing this: put the onus on the author to provide complete references or expect the journal to pay for someone else to do it.
Journal production processes are also taking measures to ensure that content is increasingly machine readable and accessible for text and data mining. Users expect to be able to mine the content in friendly formats.
Lastly, one could argue that typesetting is a relic of print; however, PDF usage of scholarly content remains the most used format by readers, though HTML full-text usage has been steadily increasing. User surveys have shown that readers like both versions: PDF to grab and stick in a file folder for reading later and HTML for search and discovery. Producing only one format would be less time consuming and would reduce expenses. That said, eliminating the PDF is not something most publishers are willing to do when over half of the usage comes in the form of PDFs.
Publishers and digital platform vendors are investing in new functionality such as allowing users to click on a figure to access the code and the data used to created that figure. This is increasingly useful to users. Inline reference tagging, high-quality recommendation widgets, annotating, and sharing features are also highly rated by end users. There is no shortage of bells and whistles that users want. However, I always go back to ask — will libraries and institutions pay for this? Will authors using grant money for APCs agree that this functionality is worth what it will cost them?
More and more, when we look at adding functionality to our products, I have to ask:
Are we considering this because it’s the right thing to do and the cost of doing business?
Do we think we can increase prices to offset the new expense?
- Interactive equations
- Data hosting
- Embedded code
Do we need to do this in order to remain competitive?
- Data analytics tools
- Taxonomy use
- Google friendly search engine optimization (SEO)
These are not always easy questions to answer.
Publishing scholarly content is not the same as just throwing a PDF online. Richly and properly tagged metadata, taxonomies, Google Scholar friendly SEO, and portable XML are a thousand times more important than whether the article title uses initial caps consistently; however, these outputs are invisible to the end user. It is far more challenging than you would think to make sure that your content appears properly in Web of Science. It takes forever for Google Scholar to correct an error, even one for which they are responsible. PubMed is increasingly stringent about its requirements. Authors are often unaware of this work until it goes wrong.
Online is not easy. Publishers have dozens of metadata agreements with some partners receiving a feed of metadata, some creating their own metadata, and others crawling a site. For the most part, there are no payments made for this metadata. It is seen as the cost of doing business and a marketing activity.
More and more, specially trained staff are required to care for metadata. A whole week can get sucked into trying to figure out why a set of DOIs are not properly resolving. This is one of many new staffing issues for smaller publishers. Positions that did not exist 10-12 years ago are fundamental to the business. These are very specialized skill sets revolving mostly around technology issues. In other words, this is not your Grandma’s editorial office anymore.
All Publishers Are Not Equal
This is where we start to see a fundamental shift in the financial paradigms between large commercial publishers and smaller commercial or not-for-profit publishers. Downward pressure on pricing across the board will not be evenly applied. There are a few cautionary notes to consider:
- Large scale publishers will be able to negotiate lower prices from vendors based on the volume of work while smaller publishers struggle to get attention from vendors.
- While large publishers have the resources to manage multiple “offshore” vendors, smaller publishers may require an “in country” vendor to help manage offshore partners. This adds to the expense.
- In some cases, the larger publishers are cutting back some services, deprecating the level of service available. For example, some publishers are no longer doing human copyediting. This leads to vendors not having copyediting staff, which eventually forces everyone to forgo copyediting.
- Publishers wishing to continue to offer services that larger publishers have forgone are pushed to boutique service providers at a much higher cost. There are fewer and fewer of these vendors.
- Commercial publishers are now purchasing the previously publisher agnostic vendors.
- Capping APCs in the OA market is worrisome to small independent publishers. Again, the large publishers have the benefit of economies of scale and can accept a lower APC than the smaller publishers can.
Here is where my big warning comes in — it does not benefit anyone, particularly the libraries, to have the entire journals market owned by 3 or 4 commercial publishers. Well, I guess it benefits the families and shareholders that own those publishers, but if you aren’t one of them, this would be a troubling reality.
All of this leaves me with some fundamental questions:
- Have the expectations for what publishers do changed?
- Are there things we still do that aren’t valuable anymore?
- Are there too many journals? Too many papers?
- What if a flip to OA costs more than the subscription model?
- If institutions are paying for OA almost exclusively, will OA journals and platforms be expected to meet the same standards that subscription journals do (COUNTER Compliance, archiving, metadata reporting, holdings reports, etc.)?
There are costs, everyone knows that. What I hoped to do here was highlight some of the activities that I would put in the bucket of “publisher added value” to potentially start conversations with all stakeholders in earnest.
I recognize that not all publishers are providing the same level of service. In my experience, self-published and society journals, or those managed by not-for-profit publishers, or even smaller for-profit publishers, have tried to use “quality of the finished product” as a competitive selling point. If those who are reading and paying for the costs of those services would prefer to have rock bottom prices in exchange for nothing but light peer review and a curated list of PDFs, some other entity will need to pick up the slack, for the sake of public safety and the integrity of science.
So I leave you with this thought that maybe, at this point, we aren’t asking the right question. Maybe instead of arguing what the cost should be — or who should pay for it – we should come to an agreement on what “it” actually is.