By John Sherer

Last Fall, consultants from Ithaka S&R visited the University of North Carolina Press to gather data they would use in writing a report on the costs of publishing a scholarly monograph. At the time, I couldn’t help but wonder whether the Press staff felt like they were being interviewed by the Bobs from “Office Space.” We were being asked how much time we spend on individual projects? How do we allocate our days? What work do we perform in-house versus outsourcing? And we were being told we would be given tools to measure our productivity and costs against our peers.

In February Ithaka released their study. No PC-Load Letter printers appear to have been harmed in the process.

Here’s what’s great about the report. It reveals in granular detail the amount of care and talent required to produce a high quality humanities monograph. And it isn’t cheap. The costs range from a baseline number of around $25,000 per book to figures three and four times that amount. By some estimates, American university presses produce upwards of several thousand monographs a year. A quick calculation suggests that UPs are covering a minimum of $50 million in expenses to make this scholarship available. I can make the argument it’s twice that amount.

But here’s what gives me pause about the report. It’s not about the data itself, which is very illuminating; but it’s about how people might use it. There’s some speculation that this report was commissioned as a step toward developing a pre-funded or supply-side funded model for publishing humanities monographs in digital open access. I certainly support efforts to rethink the economic model for publishing, especially ones where more pre-funding can be paired with a mandate to make them more broadly accessible. The paywall model has been under duress for decades, and is now almost completely broken. But with a number like $25,000 being tossed around as the equivalent of a monograph processing charge, the conversation about open access for books may end before it begins.

I agree with the consensus forming around this report that these books cost more to publish than people generally think. But I disagree with the notion that per-book subventions to presses should be in amounts like this.

The danger with the numbers in this report is that they describe how much it costs presses to put a book into the marketplace using our conventional model. But in order to produce an edition that is openly available in digital format, our activities would look very different. Or they should look very different. Any new funding model for publishing humanities monographs in open access must be paired with markedly different workflow and dissemination models. Fully subsidizing existing practices will cement existing practices, and may well introduce moral hazards: if a press is made financially whole upon the arrival of a first draft of a manuscript, our incentives to improve quality and maximize dissemination are seriously eroded. It could lead to business practices where volume and haste replace the care and quality which currently characterizes much of the output of university presses.

The encouraging news is that a digital first, open access dissemination model should be much less expensive for presses to utilize than what they’re currently doing. I’m looking forward to an analysis that reveals those true costs, as well as the expenses of shifting university presses to such a model. But those numbers are outside the scope of the Ithaka report. So I encourage you to read it, but use it as a tool to begin a larger conversation about what it is the scholarly communications ecosystem wants from its university press partners. Once we’ve developed that understanding, we can start assigning price tags.

John Sherer

John Sherer is the Spangler Family Director of the University of North Carolina Press. His career in publishing includes having been at Basic Books (where he was Publisher), Henry Holt, and the Brookings Institution.

Comments (18)

  1. Can you share the source of your encouraging news (final para)? In my experience digital first publishing costs are no lower than analogue first costs.

    1. Savings are dependent on whether you’ll still insist on creating print editions and doing cost recovery. If you choose not to, then publishers needn’t pay for composition, cover design, many (but not all) marketing activities, nor the entire bureaucracy of cost recovery (sales department, parts of finance, etc.). My argument would be that generating scholarly content and making it widely discoverable and accessible in digital formats is activity that should be subsidized. But tasks like designing covers, storing copies in warehouses, running ads in journals shouldn’t be.

      1. John, as I tweeted, I agree with your argument that the costs gathered in the Ithaka S+R report shouldn’t be used to reify the existing workflows or cost models. As you and I have discussed, there are surely creative approaches that publishers can take to rethink publishing services in an online and potentially open access environment.

        I would like to clarify that some of the specific costs you mention in the comment here – such as sales department and warehousing – were not incorporated into the cost model in the report. I don’t believe such print-only costs were even gathered in the course of the project.

        1. Thanks for your comment and clarification Roger. Indeed, the study essentially describes what we call “first copy costs” which is everything up to the moment you start manufacturing books.

          But it did include staff time from marketing and design (Table 5) in Basic Costs. Most of the activities of those employees are ultimately about the creation and selling of physical products. If you weren’t making and selling physical books, many of those costs would go away. So while an average figure like $25k doesn’t include printing, shipping, and storing books, it does include the significant amount of preparatory work that goes into laying the groundwork for physical dissemination and marketing.

          And even in a department like acquisitions, editors admit that they spend a significant part of their time doing marketplace assessments, creating pre-publication p&ls, sitting in sales meetings, working with authors on cover designs, etc.

      2. By all means consider that making scholarly content accessible in a digital format should be subsidised but my question isn’t about the method of funding. It concerns the cost of publishing content digital first. You suggest in your piece that there is evidence that this cost ‘should be much less’ than for traditional publishing. I’d love to see this evidence because in my experience (and from what I hear from other digital publishers) the cost isn’t ‘much less’.

  2. […] [This post was originally published at In the Open.] […]

  3. I agree with Toby. The experience I had at Penn State Press where we set up an Office of Digital Scholarly Publishing jointly with the Libraries and ran an OA monograph series in Romance Studies was that overall costs for OA publishing were about the same as for traditional print publishing. Yes, we did include a POD option (which of course did not involve warehousing or shipping, because the POD copies were drop-shipped from LightningSource directly to the customer), but that was a minor cost since, once set up, it operated in an almost automatic way involving no Press staff time. As for design, interior design would still be needed. Yes, some savings could be achieved by having no cover designed, but again not a major expense. A monograph series could have a template design that would require original design work only once (as with a journal), with all books past the first slotted into the format. I do not see any really major savings coming from the switch from print to digital. Indeed, if digital publishing is to do anything more than settle for being a digital analogue of print and try exploiting the full range of digital possibilities as the Gutenberg-e and ACLS Humanities Book projects did, then costs for digital publishing will vastly exceed print only costs, as anyone who has studied the costs of those ebook enterprises will know. (See, e.g., Kate Wittenberg’s post-mortem analysis of Gutenberg-e.)

  4. By the way, we ran a slide rule over our costs last year so we could quote a ‘gold open access’ APC fee for funders who wished to have books they fund free to download. We ignored all costs associated with print (including warehousing etc) and order processing etc. Editorial, production, promotion and digital dissemination costs were included. We ended up with an APC of 13,500 euros (15,200 USD) without composition (we quote for that separately since the cost varies according the number of pages).

    1. Thanks for that data point, Toby. $15,200 already sounds much better than the median numbers in the Ithaka report, which are more than twice that.

      To your other query, I don’t claim to have evidence on digital first running cheaper. My post, in fact, asks for someone to do that research so that we’re not referring to decades-old examples or simply speculating. I probably could have used the phrase “digital only”, but I didn’t want to imply that print can’t coexist with OA digital.

      However, common sense tells me that the costs to produce OA digital-only are significantly less than producing paywalled digital and print.

      And while I understand that you weren’t asking about funding, that’s what my post is about and so I’ll restate my thesis in light of our exchange, which is that conversations and proposals about funding OA monographs should start with an understanding of the costs of doing only that activity. The Ithaka study is an important step, but it has limitations for being an manuscript-processing-charge calculator.

      1. I think one of the challenges is in understanding what digital means. Print is relatively straightforward, well understood and doesn’t vary much across disciplines. Digital is not the same. In our case, we publish our e-books in three digital formats: PDF, e-Pub and READ (READ is something we developed to support our freemium OA model. It allows anyone to read our content for free on any device, including mobiles. It is shareable and embeddable. Our PDF and e-Pub editions are ‘premium’ and is how we earn the revenue to cover our costs, unless we get an APC in which case the PDF is free to download). Our books contain many charts and tables, so we offer the data as Excel spreadsheets as a downloadable file – we can have anything from 50-100 files per work. We upload each and every chapter, table and chart as separate digital objects, each with their own DOI, to improve discoverability and usability (most readers of our books don’t want to download the whole work). Typically, where we used to transform a manuscript into a single printed work, we now transform a manuscript into perhaps as many as 400 different files, each of which has to be created and managed so it appears at the right place on our publishing platform. I’m sure what we do won’t be suitable for, say, a humanities publisher, and a publisher of, say, chemistry works might have other challenges in terms of dynamic molecular models. All this to say that a digital work can be way more complex to publish than a print work (and therefore no cheaper) but it may not be the same as another digital work and therefore the costs of publishing may vary significantly across disciplines. If you want to see how we publish online, see this example http://www.oecd-ilibrary.org/development/development-co-operation-report_20747721

        1. Mr Green is quite right about the costs of producing digital publications and, as someone who works in the production, I was astounded when I read Mr Sherer’s comment: ‘If you choose not to [print], then publishers needn’t pay for composition, cover design …’. All production now uses mark-up to produce files that can be displayed -composed – in one medium or another; it has done so since we stopped using hot lead and Rubylith. This mark-up must be applied correctly so it can be interpreted by the rendering engines, whether this be at the print shop or on a personal device or a website. In addition, PDF is by definition paginated so its pages must be composed just like those destined for print. And, EPUB has strict rules about its mark-up and, because support for the standard varies greatly from device to device and is often defective, publishers must devote resources to what is euphemistically called ‘quality assurance’. If one publishes with Amazon, the digital document must be converted into yet another format and its display again tested on the targeted devices. Finally, the standards for digital formats require a cover that will appear as a thumbnail on the device or the website from which the document is distributed.

    2. Did these costs include any overhead, Toby, or were they all direct costs? If the latter, that could help explain the difference between your cost and Ithaka’s.

      1. As ever, please define overhead 😉. However in our case it excludes office space and services like human resources but it includes pretty much everything else.

        1. That is indeed a problem for university presses because each press has different overhead expenses to account for. E.g., at Penn State the Press was not charged for any cost of staff benefits, so we didn’t need to include this cost in our calculation of overhead. Thus using what any one press does as a basis for coming up with a model for all is going to be suspect. So, did your overhead allocation include a proportion of staff benefit expenses? Utilities? Legal fees?

          1. Our costs exclude housing & utilities, HR support, accounting support (since these are provided by the institution). Any complex legal stuff is also handled by the centre, but day-to-day issues by us (legal stuff is a tiny part of our work). All staff costs (editorial, marketing, production, IT, admin etc) are inside our budget. We get a financial contribution from the institution (an amount that is roughly 10% of our budget, but it declines in real terms each year) to help cover some of the costs of being part of an IGO, such as running a publishing programme in languages other than English. (The harder one looks, the more we learn that each institution is unique making comparison, as you put it, ‘suspect’!)

  5. The University of California Press has already put a price tag on OA mongraph publishing: $15,000 from the author’s institution plus $7,500 from the author themselves. Not too far off from your $25,000 estimate. http://www.luminosoa.org/site/faqs/

    1. Anna, I think Luminos is $15,000 total, rather than $15,000 PLUS $7,500. However that number is based on a number of business assumptions at UC Press (e.g., whether there will be continuing revenue from print editions, how big a part of the total operation Luminos will be) and I do not believe they claim that it is intended to reflect the actual cost of production of an OA monograph.

  6. […] a possible transition to new publishing models at traditional presses. John Sherer has argued that “Any new funding model for publishing humanities monographs in open access must be paired with mar… making clear that costs should not be seen as fixed. Strategic issues on the supply side thus may […]

Leave a Reply to Roger Schonfeld Cancel reply

Your email address will not be published. Required fields are marked *