Editor’s Note: Today’s post is by John Sherer. John is the Spangler Family Director of the University of North Carolina Press. He is the chair of the Association of University Presses Open Access Committee and is the Primary Investigator in the Andrew W. Mellon Foundation funded, Sustainable History Monograph Pilot.

Within the scholarly book publishing community, it’s not particularly controversial to claim that free digital editions of monographs will erode print sales. After all, who would pay for something they can get for free? These books already sell so few copies, and the economics are so unfavorable, that further revenue erosion could easily shatter an already precarious ecosystem. That said, there’s a growing body of research indicating that readers strongly prefer print formats for these publications (for example see the 2018 Ithaka S+R Faculty Survey and Naomi Baron’s Words on Screen). And there’s anecdotal reporting that in open access (OA) experiments at university presses, print sales have been stable. Can we review sales data for OA titles to find out if the claim of print cannibalization is true?

E-book reader, stack of books

The National Endowment for the Humanities has just funded a research project to empirically review whether the availability of OA editions of university press monographs has a quantifiable effect on the sales performance of print editions. I’ll be working with my co-PI, Erich van Rijn from the University of California Press (who helped draft this post), along with the Association of University Presses (AUPresses) and Roger Schonfeld (full disclosure, a regular contributor to this site) and Laura Brown from Ithaka S+R to collect sales data and try to draw some conclusions.

Are Monographs Too Specialized or Too Expensive?

The collective work of university presses to publish and distribute monographs is one of the cornerstones of the creation and advancement of scholarship in the humanities and many of the social sciences. By at least one estimate, more than 4,000 new monographs are published by university presses annually. While sales have been in decline for many decades, the last decade’s development of digital distribution networks created the potential for dramatically reversing this trend and increasing the accessibility, use, and impact of these monographs. But even though the vast majority of new university press monograph are published simultaneously in print and digital formats, there has not been a dramatic increase in revenues. There has not been a major expansion of collections development by academic libraries, despite significantly lower per-title costs. Does that validate a common criticism that these books are simply too specialized? Or are the obstacles still financial? If you could remove the paywalls to digital editions, would usage and impact significantly increase?

University Presses and the Marginalization OA

There have been a number of university press OA pilots in the past five years including NEH/Mellon Open Book grants, TOME (Toward an Open Monograph Ecosystem), Knowledge Unlatched, NEH Fellowship Open Book Program, and the Sustainable History Monograph Pilot. In addition to some exciting new start-ups, established presses like MIT, Michigan, and California have also shown extensive leadership in opening their own books. Foundations have contributed millions of dollars to initiate pilots and support capacity-building for open models of publications. Through these and other initiatives, there are now thousands of university press books that have been published in open digital editions with print copies for sale.

While there is a prevailing desire among membership within AUPresses to do more OA publishing, in internal surveys we’ve confirmed that presses cite long-term, sustainable funding models as the top problem in implementing OA. Even with the millions of dollars spent in experimentation, OA remains very much on the fringes within the university press world, making up less than 5% of new monograph output. In a permanently challenging fiscal environment, the prevalent assumption that open digital editions will cannibalize revenues has stopped OA from growing, resulting in the suppression of a substantial body of humanities scholarship. However, no one is citing data on print cannibalization. It just feels like common sense.

And then the COVID-19 pandemic created an unwanted opportunity to conduct a test.

The Unexpected Pandemic Experiment

During the early months of the COVID-19 pandemic, when students, faculty, scholars, and researchers were being displaced from campuses and research centers, platforms like Books at JSTOR, EBSCO, and Project MUSE asked university presses to temporarily make their digital monographs free to read and open to anyone via so-called emergency libraries. Not every press said yes, but many did. This created an unprecedented real-world experiment where these vendors could look at the usage of books that had long been paywalled but were suddenly free to read. No one was surprised that usage went up. But many of us were surprised by how much their usage increased and that much of the growth occurred in regions which have historically lacked access to such works. At the same time, many presses saw increases in print activity. It was a unique moment and we’ll probably never be able to unpack everything that was going on, but during those months in the spring and summer of 2020, when much of the world seemed to be falling apart at the seams, the myth that university press monographs were too esoteric to be read was shattered. These books circulated in both digital and print formats at unprecedented rates.

What We Hope to Learn

Even if it is true that in more stable situations, OA erodes print revenues, then understanding the degree to which print is reduced could be a key guidepost in modeling a sustainable OA subsidy. But if it is not correct—if OA has only a marginal impact on print revenues—then there could be a substantial opportunity to expand OA publishing. But let’s also give ourselves permission to imagine a more surprising result. What if the discoverability afforded by OA leads to an increase in print revenues? What if highly accessible digital content becomes a marketing tool that leads to print transactions? Could there be a scenario where the global impact of university press monographs increases exponentially, and print revenues ride those coattails? It could have a transformational impact on the future of monograph publishing and the accessibility of humanities scholarship writ large.

John Sherer

John Sherer is the Spangler Family Director of the University of North Carolina Press. He is the chair of the Association of University Presses Open Access Committee and is the Primary Investigator in the Andrew W. Mellon Foundation funded, Sustainable History Monograph Pilot.


48 Thoughts on "Guest Post — Does Open Access Cannibalize Print Sales for Monographs?"

Although monograph is intellectual property,it should cost something but sometime many new scholars who really wants to gather authentic and pure knowledge and at the same time unable to pay them OA is the only way to nurture themselves.
As I suggest there should be a condition before providing OA like why and who wants monograph and publishers can keep track on that.As almost every reputed publishers do.

Pretty much everything going on in scholarly publishing trends cannibalizes print monograph sales. And that has been going on for a long time now. Academic libraries consistently report lower and lower spending on print monographs every year for over a decade now. Just-in-case purchasing of print monographs, which was the only option for over a century, is now increasingly seen as the really lousy deal that it is for most libraries, now that better purchase/license models are available. Publishers need to know that librarians are increasingly understanding this, and they need to stop bemoaning the drop in buggy-whip sales and start taking seriously their own need to change their revenue models. It’s time to think big – this goes way beyond whether this or that book is OA.

This is going to be a tough study to interpret. During the pandemic, many academic libraries weren’t purchasing any print monographs because no one was in the library to physically process the materials. If you are able to conclude that there is evidence that OA cannibalizes print sales, it won’t be very generalizable, unfortunately, only that OA cannibalizes print during a global pandemic.

Phil–I agree completely. We did an analysis of the effect of print on OA during the pandemic for Johns Hopkins University Press titles. We made all Hopkins Press titles on MUSE OA during the beginning of the pandemic. When we talk about that analysis, we are very careful to communicate that we cannot generalize our results. To generalize results, we need to conduct a scientific, randomized experiment.

I missed this comment earlier. I just want to signal my full concurrence.

Hi Phil: I didn’t make space in the post to explain this, but our goal is to create a data set and methodology that would allow it to be looked at it from different perspectives and over time. We’ll offer our own analysis as well, but will look forward to multiple interpretations. But the reason we started this project was to overcome the default assumption that OA hurts print. That seems to be taken as gospel in every conversation. Even if all we do is muddy that presumption, it could be progress.

Is there any data at all on the mentioned “increase in print activity” during the pandemic. That had not been the experience of any of the scholarly publishers with whom I have talked, so I am wondering if it applies to particular subject areas. Or perhaps university press trade books?

I remember seeing in the last year, although I can’t find it now, that a major ebook platform seller (I think it was EBSCO/Gobi, IIRC) reported that for the first time, they saw ebook sales exceed print book sales, during the pandemic. This would be a more reliable report of what libraries (who are the main purchasers of scholarly monographs) are actually doing. More popular-press reports about print and ebook sales (and preferences for that matter) are typically using data driven by individual reader sales, like Kindle books, and other mass-market industry data, which is not at all the market for university press books.

John, just to clarify — Lynne’s question was about a reported “increase in print activity,” but your response addresses “university press sales,” which isn’t the same thing. Do you have data on the impact of COVID on sales of print, specifically?

I’m afraid I may have gotten lost here, Rick. Help me out.

In my original post I wrote “At the same time, many presses saw increases in print activity.” Should I have been more clear that “presses” meant university presses? Was that not what Lynne was referring to? The data behind that statement is from the AUPresses internal sales tracking. Although the AAP data does suggest sales went up for all kinds of publishers during the pandemic.

To answer your question about whether I have data on the impact of COVID on sales of print, specifically, that’s a hard no. OTOH, most publishers rarely know what drives sales even during the calmest of times. Too many intermediaries in between us and our readers. So all I was trying to say is that during this great experiment when we opened thousands of books, one might have presumed that print sales would crater. And they did not. All evidence suggests they increased.

Here at the Oxford International Centre for Publishing, we have published a report on monograph publishing, using data from publishers. The report covers a number of aspects including the number of titles, revenues, print runs, subject areas and digital developments.

One respondent to the survey commented on the future of monograph publication: ‘it is likely to look dramatically different. We anticipate that in ten years’ time a much larger proportion of the monograph list will be published open access, and that new models will be necessary to manage this transition in a sustainable way.’


When you list “university press OA pilots in the past five years” don’t forget Lever Press (https://www.leverpress.org/). Funded by liberal arts college libraries, digital first, always open access, where a print copy can be purchased, but really, why bother?

Thanks for mentioning Lever, Jonathan. Another factor in this conversation is the declining demand for print monograph titles that libraries are experiencing. Overall circulation of print titles declines every year, and that of course well predates COVID. To Melissa’s point above, not only are budgets being cannibalized, but librarians need to consider whether the demand is there for building large collections of print titles, whether we can afford them or not.

Thank you Jonathan. I had to use the phrase “in addition to some exciting new start-ups” to cover folks like Lever/Amherst College Press and many other terrific initiatives. But I also wanted to make the point that it is legacy university presses that tend to use the cannibalizaiton argument as a reason to not try OA.

Hi Jonathan,

I did elide over this with my comment about “some exciting new start-ups” but I appreciate your mentioning this. The argument that OA cannibalizes sales is coming from legacy presses so that’s where my focus was. But we’ll be relying on the data from start-ups like Lever/Amherst College Press.

I would be very interested in learning more about the operating data for Lever Press. How many books published to date? How many are in the pipeline? What is the cost per publication? What problems have arisen/what strengths does Lever propose to build on? Where does Lever see itself (with quantifiable data) in five years? As for the sarcastic “why bother” comment, part of the problem with so many experimental start-ups is that they communicate nothing. One has to know all about them in order to find them in the first place.

There’s a fair amount of information on Lever’s website, including links in the news section to our periodic member newsletters: https://www.leverpress.org/. But yes, we could certainly do a better job of communicating out where we’re at with Lever’s progress, and where we see it going in the future.

Interesting thoughts John—thanks. In the interest of looking for broader lessons and patterns, maybe it would also be worth looking at what we know about the finding the right balance between free and paid content in journalism (https://bit.ly/3rYS1pN), tech (https://bit.ly/3u9UIaT), movies, music (Rick is an expert here), and so on? In journalism anyway, the short answer isn’t either/or. Increasing the right kind and right amount of free content to the right audiences can indeed result in an uptick in revenues (where ad revenues are generated from those with a low willingness to pay and subscriptions are enticed from those with a high willingness to pay).

At my library, we would like to have more opportunities to invest in open monographs. And we would like the UP move to open models to happen sooner rather than later. While we will continue to support paywalled models like the De Gruyter UP partnership program in the short term, this type of investment is of diminishing value against spending that expands equity by opening content to the world. I don’t believe UPs should consider closed business models a safe, long-term harbor.

Hopefully this new study will help establish open models as a financially viable thing to do. We already know readers are waiting and that, for equity, it is the right thing to do.

To add to the dilemma faced by libraries, we are forced to choose between print and e when making title-by-title selections. The fact is, both formats have value. I made a case for publisher bundling of P+E monographs in a recent edition of Against the Grain (https://issuu.com/against-the-grain/docs/atg_v33-4/s/13538493) and will be leading a Table Talk on the subject at the upcoming Timberline Institute.

Librarians aren’t “forced” by the publishers/vendors, just their own budgets. Lots of publishers are offering steep discounts for us to purchase ebooks of books that we have the print of already. My own library just can’t afford to spend money twice on the same intellectual content when there is so much else that we don’t have and would find useful in our collection, and we have a general policy (with some specific exceptions) that we do NOT purchase both print and e for any title but our subject librarians get to choose when making title-by-title firm purchases.

The argument in my article is that (1) you shouldn’t have to pay twice for the same content, and (2) you should have the opportunity to purchase both formats, at a steep discount, at the time of publication. The “sales” publishers offer on ebook versions of owned print titles are inefficient from an acquisitions standpoint, and they don’t provide immediate access to both formats.

You have my 100% support on the first point, but copyright law grants monopoly powers to the copyright holder, so our “shoulds” mean absolutely nothing. We librarians either need to get the copyright law changed, or find a way to create a revenue-budget win-win for publishers and us. The publishers are in a zero-sum game against each other for our few remaining monograph dollars. It’s clear to me from numerous conversations with publishers (I’m an acq librarian and talk to a LOT of major publisher sales reps, almost every week) that they understand that. But it seems a lot of the non-profit small presses, esp. university presses, really don’t, and they suffer from a massive delusion that THEIR books are just so wonderful we librarians won’t be able to resist buying them just-in-case. We will, we have to. It’s time they stop living in 1985 and get with reality.

It’s worth noting that most books, including academic books (and, yes, university press books) are sold to individuals, not to libraries. Libraries are increasingly irrelevant for the financial performance of book publishers. A strategy, whether OA or a traditional one, based on libraries’ centrality is not forward-looking. Please visit the home of a professor in the humanities and count the bookshelves.

Perhaps this varies by discipline. And it would be interesting to see if that “most” is by volume or by dollar, given the huge markup libraries pay for ebooks compared to individuals and to print books. Btw, in an earlier library job in my career, I was made aware of a humanities professor who spent a lot of money getting massive custom bookcases built in his home. But the books he put on them were predominantly from our library, which had a very liberal policy of letting faculty just keep renewing their checkouts semester by semester, in bulk, without having to bring any of them physically in and without having to respond to “recalls”. That one prof had so many of our books on what we all decided was “permanent loan” that they made a big hole in one area of the LC stacks and we ended up just shifting into it, deciding we didn’t actually need to keep those shelves empty because we weren’t going to see those books back for a very very long time.

The “most” is by dollars, not by units. Retail customers subsidize libraries, which in turn cannibalize retail sales. Pricing should be by use, not by copy.

I don’t disagree, Joe, that revenues from libraries are becoming a smaller fraction of university press income. But I do think they’re a key player within a series of discoverability steps that leads to a consumer sale. Many scholars initially find a UP monograph through the discoverability and reading tools at places like MUSE/JSTOR/EBSCO. It turns out monographs aren’t much fun to read in web browsers, but that’s where many people find out about them. But if we had to compete on the open web for eyeballs, we’d lose. So the way libraries surface UP content is vital to our sales. Even if they’re not the ones giving us the most money.

John, isn’t the really a comment on what a poor job publishers do in marketing their books? Web sites are hard to navigate (and take forever to load), metadata is incomplete, SEO is weak to nonexistent–and on top of this many publishers encourage readers to read PDFs! It’s definitely true that libraries are part of the ecosystem, but they are a peculiarly privileged part, and book publishing is a notoriously difficult place to make a dollar.

Thanks for your comment, Joe. I mis-wrote when I said that these platforms is where “people find out about” our books. Some readers probably do, but what I was intending to say was that publishers’ marketing to scholars at institutions with strong libraries drives an electronic search that frequently takes them to a library reading platform like MUSE/JSTOR/EBSCO. Or Amazon, if they’re ready to buy. But the former if they want to do a “reading dive” to see what’s in the book. My OA argument here is that if at this stage of discoverability–the time a reader takes to dip into an electronic text–if we expands access rather than paywall and delimit it, then we might drive more print sales in the long run. At least that’s they hypothesis we’re testing.

I’m really curious what your criteria are for defining a “strong library”.

Melissa–can’t seem to make the blog let me reply to your reply, but my shorthand for “strong library” was one that had the ability to already buy wide a swathe UP monographs for their patrons.

I can’t agree, Joe, that university libraries are not important for publishers of scholarly, peer-reviewed monographs and collections. Before the pandemic, those libraries accounted for more than half of our sales. That has dropped to less than 30 percent during these last two difficult years, true, but that still represents a significant audience.

I think your organization may be an outlier. The last time I looked into this (pre-pandemic, as part of a Mellon project) libraries accounted for under 20% of American U. press revenue, a figure that was trending down. Interestingly, the same amount was derived from traceable course adoptions. Admittedly, the U. press world is not a microcosm of academic books overall, but directionally I think these figures are important for strategic planning. This is why, in my view, the emphasis on OA for monographs is a distraction from the bigger opportunity in D2C marketing of books. YMMV.

From a social science perspective: The use-value of a well-prepared and reviewed monograph/book extends often beyond the author’s lifetime. This is something that rarely happens in the journal world, where quick results, fancy headlines are currently sought after. We should be therefore careful in extending our biases towards or against OA, stemming mainly from the experiences in the journal world, to book performance in general.

I would really recommend anybody who hasn’t yet looked at the big OBU report flagged by Angus Phillips above to do so: this makes a brutal and under-articulated truth starkly clear, that whilst university presses are very important to global anglophone monograph publishing in the arts and social sciences, they are not of themselves determinant. Commercial imprints large and small continue to publish the significant majority of AHSS monographs (perhaps especially in social science subjects) and any discussion about ‘monograph futures’ which doesn’t recognise that is missing something. This new AUP initiative is really welcome, and should provide important evidence in a domain in which (at present) anecdata is rather too prominent (not least amongst policy-makers). Will John and his team be collecting global or at least international data or largely restricting their enquiries to North America? It would be a real shame if it was just the latter…anyway good luck to all concerned!

I launched Business Expert Press in 2008 with a multi-channel/multi-model (including open) distribution strategy from day one. The more digital chanels we opened (with as many partner per channel as possible), the more price points, the more access models, and the more open we did the more total revenue we generated per title. Period. Please do your study. But I am doubtful it will erode the “conventional wisdom” that sees the consumer as a rational alternative seeker (they are not). This “fixed pie,” “conventional wisdom” perspective that has led many a senior publishing executive to steer away from distribution experimentation on the premise of protecting existing revenue models (e.g. print) drove nearly all publishing genres into revenue decline just as digital opened up endless new avenues to growth. Don’t blame declining library budgets, and don’t blame declining interst in reading; blame conservatism and timidity across the leadership of the publishing world.

I have made this statement before, but I’ll repeat it here. When I started in university press publishing, we published for libraries. I don’t anymore. Now I publish primarily with Amazon in mind. Libraries haven’t been a significant market for university presses for a decade so I stopped catering to them years ago. And my press has a very aggressive OA policy. ALL of our monographs become OA five years after publication. You can find them here. https://kb.osu.edu/handle/1811/131

So, what was being bought? The biggest increase in ebooks came from libraries that were replacing print copies with digital copies at a huge discount, but what surprised me was they were buying the backlist that they could have gotten for free at OSU’s IR. And print and e was still selling at Amazon and those sales increased significantly. The question I have is who was buying that print and e? Amazon will never tell me but if I had to guess I’d say it was faculty and grad students who found Amazon a safer and more expedient choice than their own library. It probably helps that we really strive to publish affordable paperbacks, rarely going over $30.

University presses aren’t monograph monoliths. Our trade program also sold quite a few books and they absolutely help our bottom line. But the biggest surprise to me was our textbook. It’s one of the most used intro to linguistics texts in the world and can provide up to a third of our budget during the first year of sale for a new edition. But we decided to open it completely on OSU’s IR at the beginning of the pandemic and what happened after that stunned me. It was open for six months and in that time it had 41,894 downloads of the DRM free PDF. I thought for sure it would mean the collapse of sales for that book, instead sales increased not just in FY 2021, but to this day. This is a six year old edition and sales continue to increase.

There is so much we don’t know about the impact of open on consumer behavior, and I for one welcome this study and hope we learn a little more about this subject through its results.

Hi, so I went to your site to find this textbook you described, and it appears to be locked back up again.
all of the download links say “admin access only” and lead to authentication prompts.
Aside from my disappointment about the one book, it does bring up issues to do with OA that we librarians worry about and may explain why you continue to see library sales of books that are “free”. We librarians aren’t yet at the point where we trust the long-term reliability of OA, and we librarians are all about building collections in the long term.

It’s like I tell people who ask why I still buy music on CD: I love the cloud, but I don’t trust it as a permanent source of access.

As I said in my comment, we opened it for six months. For the rest of the spring semester and the summer semester. It is not permanently open but the PDF files that were downloaded are still usable. The only other book on OSU’s institutional repository that we subsequently closed was a trade book that my predecessor actually added to the IR. These are not in library catalogs because there are so few catalogers left in higher ed libraries, and of those, so few tasked with adding OA to library catalogs, which is why faculty and grad students go to Amazon where books are far more easy to find.

If the global context is of interest, the UCL Press statistics may be of interest. The over 200 titles are mainly (but not exclusively) AHSS, published OA (available via UCLP and now JSTOR etc) but also available in paperback and hardback. Titles include both monographs and edited volumes. Broad-brush download plus sales figures are published for each title monthly (https://www.uclpress.co.uk/pages/statistics ). Does OA preclude hard-copy sales? No, but it’s complicated. A volume of essays I co-edited, The East India Company at Home, has consistently been in the ‘top ten’ downloaded titles for the past year or so, as well as in the ‘top 10’ print sales (https://www.uclpress.co.uk/pages/statistics : search under ‘The East…’: as of 1 Feb 2022, 90,023 downloads and 544 print sales, N=222 titles published). It appears that in some cases OA availability drives purchase of hard copy (especially, I think, for image-rich texts). But disentangling correlation and causation is tricky, and deserves much more systematic attention, not just for monographs but also for edited volumes (the latter seeing much use for undergraduate teaching in AHSS subjects). Really excellent to see this new research project under way.

Thanks for your comment, Margot. We will definitely be tapping into the experiences at places like UCL Press who have been real leaders in the born-OA university press movement.

Let’s do some A-B testing. If OA drives print sales, then what would happen if only half the book were OA? One-quarter? 10%? Two pages?

I appreciate the responses to my brief comment. I was interested in “unpacking” the reported “increase in print activity” during the worst of the pandemic, because, as I mentioned, that was (and is) not LRP’s experience nor that of my colleagues in scholarly publishing. I was wondering whether it was the trade-oriented books that were doing so well, or the specialized monographs too. If the latter, I will be both surprised and humbled.

Comments are closed.