University of Michigan Press is part of Michigan Publishing, U-M’s academic publishing division, and part of the University Library. Michigan Publishing recently celebrated its ten-year anniversary as the umbrella for all of the library’s publishing activities, but its history with open access (OA) publishing stretches back even further. The Scholarly Publishing Office was started in 2000 as an exclusively OA operation, publishing small journals in the humanities, digital projects, and the Maize Books imprint. In 2006, the Press launched its first OA imprint, digitalculturebooks, and has since put out about 120 front list titles and made freely available about 800 older books through HathiTrust. This history demonstrates U-M’s long-standing commitment to openness and experimentation. I spoke with Charles Watkinson, the director of the University of Michigan Press, in January about UMP’s Open Access program and the costs of OA publishing. The interview has been edited for length and clarity.
Yelena Kalinsky: What was the impetus behind Michigan Publishing’s original move towards OA?
Charles Watkinson: I think it was fairly ill-defined at the start. It was part of a general move within the library environment towards supporting open access from a values perspective, as a public good. Over the years, that attitude has become a bit more pragmatic and a bit more focused. The values-based arguments for open access publishing remain very present, but there is also a more pragmatic conversation about who explicitly is one targeting with each different open access project that one does, and who gets the benefit. It varies by project.
The values-based arguments for open access publishing remain very present, but there is also a more pragmatic conversation about who explicitly is one targeting with each different open access project that one does, and who gets the benefit. It varies by project.
YK: And when you say project, you’re not just talking about monographs.
CW: No, it could be an open access journal, a dataset, or a digital product which doesn’t resemble a book. I am particularly interested in books because that is the majority of what the UMP does. But it could also be things like technical reports, grey literature of various sorts, all kinds of products.
YK: What portion of new books published by UMP every year are open access?
CW: About 10-15%. We plan to increase that number over the next 5 years and we are currently trying to work out what percentage we will increase that number by, how aggressive we can be. And that all depends on the funding that we have available.
YK: What is the breakdown between the different models (platinum, gold, and green)?
CW: We try very hard to veer towards platinum models (i.e. models where neither the author nor their parent organization has to pay), because we do get concerned about excluding authors who are unable to pay. I would say that at the moment, it is probably about 80% platinum open access and about 20% gold. The gold program we participate in is mainly funded by the TOME (Toward an Open Monograph Ecosystem) initiative, and in that case it is the author’s parent institution rather than the author themselves who is paying. We don’t really do a lot of green. We do have allowances in our contracts for authors to post their manuscripts online in repositories soon after publication, but we don’t actively think about green as a strategy.
...we don’t want to sacrifice the acquisitions editorial activities. Because that’s where 50% of the costs lie, that’s a particularly tricky question, how we can streamline that very labor-intensive activity without throwing the baby out with the bathwater.
YK: Can you walk me through the different costs associated with OA publishing? Are there savings associated with OA?
CW: First copy costs still exist, and those are the costs that go into producing the file from which an ebook and a print book will be generated. First copy costs for the UMP are substantially made up of the acquisitions editorial functions, the work of acquisitions editors. That includes the work of selecting books to be accepted for publication and the work of developing and improving the work that’s accepted. That is actually a very substantial part of our costs, just under 50% of the cost of the first copy. The other activities are related to production and marketing, but those are less of the total first copy cost. The challenge is that open access in itself does very little to make our costs lower. There may be some reduction in cost in the distribution, because there is less need to focus on getting books into bookstores or retailers, but the reality is that we haven’t as a specialist academic press been getting books into bookstores anyhow for a long time, so the costs there are not very large. If one thinks about the cost opportunities, they’re not about open access as such, it’s more about finding ways with digital technologies to reduce the costs of production and that’s something that we are working on.
Then there are other costs associated with digital innovation and making products available that are not simply books that are actually increasing. So we have a pressure for cost reduction conflicting with an author demand for more functionality that makes our cost situation a little bit complicated at the moment. We have done several analyses of costs, and each monograph costs us about $27,000 to produce, taking into account all costs to first copy. We do want to get that number down, but we don’t want to sacrifice the acquisitions editorial activities. Because that’s where 50% of the costs lie, that’s a particularly tricky question, how we can streamline that very labor-intensive activity without throwing the baby out with the bathwater.
YK: I know that the cost of the first copy ranges very widely. What could be the reasons for that?
CW: That was a finding of the ITHAKA study from 2016. That project did have a few outliers in it. There was a very high number of something like $130,000 and that was an art history book. The elements there were a lot of art that had to be sized perfectly, reproduced perfectly, there were some permissions clearance fees in there, so there were quite a lot of exceptional costs in that particular outlier. At the other extreme would have been a simple text-only book from an author who has done a good job of preparing their work before submitting it to the press. So, I think the thing about those outliers was that there were substantial reasons why those books were pretty different from the usual kind of monograph. The usual kind of monograph in that study, did average out at about $30,000.
... every project is different. What we're looking for with open access books is a strong argument for why open access would be a good strategy for this particular work.... When we identify open access as a good strategy ... we will subsidize the upfront costs of an open access book. It’s a pragmatic estimation in our minds and that is why we are happy to share the risk.
YK: UMP does not have a single standard book publishing charge (BPC) and often charges substantially less than the first copy cost to publish an OA monograph. Where else does funding for OA publications come from?
CW: This comes back to the idea that every project is different. What we're looking for with open access books is a strong argument for why open access would be a good strategy for this particular work. So in political science that argument might be around influencing public policy, for example. With a humanities book, you may have an author who wants public annotation to be able to happen, so the affordances of digital scholarship would be a good argument. It really depends on the project. If the project is right and the author is networked with a particular community that might actually spend money on buying a copy if they saw an open access edition, or if the market might be expanded by having an open access edition, we make a fairly pragmatic decision. When we identify open access as a good strategy (especially one that will be likely to make us as much money or even more money than we would otherwise expect) we will subsidize the upfront costs of an open access book. It’s a pragmatic estimation in our minds and that is why we are happy to share the risk.
YK: If there is a large audience for a particular OA book and readers can access it for free, why would that help with recouping costs?
CW: Well, often readers like it so much they want to read a print version. That’s how the extra sales tend to happen. Also we haven’t traditionally made our books open access for Kindle or Apple iBooks, so some readers will choose the convenience of downloading a Kindle version to their own devices and be willing to pay something for that. It’s really about the idea that open access gets this extra exposure, this extra buzz, this extra demand for the work, and then a bunch of people will access the book for free electronically. But there will be this other group who will be willing to spend some money to actually buy a print copy or a downloadable ebook, and that’s where we’re willing to subsidize, because we see an opportunity.
What we are increasingly seeing is that we have enough support from libraries that we could do more. We could put more of our money on the table to create a fund that wouldn’t require the author to actually pay anything. It’s a bit of an extension of our participation in Knowledge Unlatched, but with the idea that it’s a more generalized fund that can be used for the books that we think are valuable rather than the books that the Knowledge Unlatched librarians choose. We see open access as an extra tool in our acquisition editors’ lockers when they go after particular projects.
But if one looks at a press that’s very heavily reliant on monographs, that's where you start to see cracks emerging. And those presses are needing to turn to their parent administrations and say, look, is this truly a mission-related part of the university’s activities that you’re willing to subsidize or support as part of the core academic mission?
YK: You’ve been speaking about the mission of university presses. They aren't meant to be profitable, but we’ve seen administrators making the opposite argument. Where does OA fit into that conversation?
CW: That's a super question and very difficult to answer. What makes it hard is the diversity of different presses. How they sustain themselves and how administrations views them is usually very different. There's a real problem at the moment with monograph programs within university presses and that’s disguised somewhat by the fact that most university presses cross-subsidize their monograph programs. It might be with a regional trade program--books about the region, cook books, natural history books, that sort of thing. It could be a journals program, or a textbook program, or distribution and services for other publishers. Any of these could be an area that a press can expand to offset the losses in their monograph programs. The problem is when one looks across university presses, they all have slightly different models of cross-subsidy. But if one looks at a press that’s very heavily reliant on monographs, that's where you start to see cracks emerging. And those presses are needing to turn to their parent administrations and say, look, is this truly a mission-related part of the university’s activities that you’re willing to subsidize or support as part of the core academic mission? And the answers that they’re getting really depend a lot on the administrators who are responding and how well they know about university presses, how they view their own institutions, and their responsibility to the whole system of our education. That's why there's so much variability.
What open access is doing, which I think concerns top administrators at universities, is that it is threatening to put even more of the costs of scholarly publishing back on the universities who are the most research-productive. That's what we're seeing in the journals space with article processing charge-funded journals publishing. The institutions who produce the most scholarships are likely to bear the brunt of paying APCs in this new environment. And that’s also the danger that we see with monograph publishing. That institutions who have to pick up more and more of the costs of operating their university presses are benefiting the system broadly, but they’re bearing even more of the costs of supporting it. So I think that's a really difficult part of open access, this redistribution of who's paying. That's pretty difficult for big public research institutions like MSU or University of Michigan. Those are some difficult decisions to make. It's messy, complicated, and also very difficult to compare across.
This interview is the first in a Feeding the Elephant summer series on the scholarly publishing ecosystem. We invite readers to participate in the response box below and stay tuned for more installments!