As discussed in the Scholarly Kitchen a couple of weeks ago, the University of California (UC) Libraries have recently released the final version of the Pay It Forward report, examining the impact on large, research-intensive institutions in the US and Canada of a wholesale shift to gold open access (OA) for journals.
The search for feasible OA solutions has gained momentum over recent years. Although the most recent Outsell report estimates that gold OA has grown to a market of about $300 million, it also forecasts that this growth will slow down for lack of a robust infrastructure. At the heart of this infrastructure is researchers’ need for a clear understanding of who will pay article processing charges (APCs). Other recent papers – such as that from the Max Planck Digital Library – have argued that a large-scale “liberation” of library subscription dollars to fund gold OA would be no more expensive and perhaps even cheaper than the current system. The Pay It Forward report is the first comprehensive analysis of how a decisive push to Gold OA would impact the cost of publication for major North American research institutions.
I spoke with the UC study’s co-principal investigators, MacKenzie Smith and Ivy Anderson, about the findings of their study and how these fit into wider prospects and challenges for accelerated moves towards OA.
(MacKenzie Smith is University Librarian at the University of California, Davis, and Ivy Anderson is Director of Collection Development and Management at the California Digital Library.)
There have been a various studies of the costs of OA – can you tell us about your motivations for undertaking this particular project?
As more authors from the University of California choose to publish in OA journals funded by APCs, the UC Libraries have gotten more involved in helping with those fees and advising authors on their options. That, combined with unsustainable journal subscription cost increases, made us look seriously at the cost of converting journal publishing entirely to fee-funded OA, which we considered very worrisome. Our preliminary calculations — built on the earlier study done by Phil Davis in 2004 that David Crotty’s post mentions, but extended to include a wider range of variables such as grant-funding and co-authorship patterns — suggested the need for more analysis, so we decided to conduct a rigorous investigation of this question for institutions like ours. And not limited to today’s costs, but thinking carefully about how those costs might evolve in the future, as the majority of journals shifts to APC-funding. We also realized that attitudinal factors would play a key role and so combined qualitative and quantitative approaches to develop a richer overall model.
What do you consider to be the key findings? Did any of them surprise you?
First and most obvious, that research universities with high publication volumes will experience higher costs than today, if fees don’t go down. This confirms what we suspected but contradicts recent studies, particularly in Europe. The claim that this study just confirms what we already know is overblown. The only previous study was a simple spreadsheet that used aggregate public data and a limited set of variables. Our own, more customized version of that analysis is what motivated us to do this project — with millions of dollars and careers at stake, we need more than back-of-the-envelope calculations.
Second, while authors have wildly different attitudes about OA and paying for journals, they are sensitive to those costs when they think about different sources of funding. That isn’t surprising if you understand market economics but also contradicts the common belief that authors don’t care about publication costs at all.
Finally, we found that, despite higher overall costs to publish, there are scenarios in which research universities can afford a system of APC-funded journals, and even scenarios where costs could go down over time.
We had also hoped to be able develop a more differentiated model of APCs in different disciplines; our hypothesis was that APCs outside of the sciences would be demonstrably lower than in STEM fields. Unfortunately our study didn’t surface enough statistically valid data to back that up, although there are some indications that this sort of differentiation may evolve over time.
The report confirms earlier conclusions that high-intensity research institutions – such as the UC – would shoulder a much higher burden of publication costs in a Gold OA model. In a world of flat or declining funding for higher education, what factors do you believe will persuade institutions to move in this direction?
This is an excellent question. What we found is that, for top research institutions, library budgets alone are not enough, if APCs stay at their current levels or go up. However, if granting agencies begin to play a larger role in funding publication, most institutions would not have to shoulder a greater burden. For all but two institutions we worked with, if research grants cover APCs when they can, there is more than enough library funding to cover the rest; and the gap for the remaining two institutions was not large.
With respect to incentives, the greater value that OA offers in terms of visibility, utility, and research impact are powerful motivating factors in a world where the health of our universities increasingly depends on the ability to demonstrate the value of the research enterprise to society at large.
The second significant barrier would seem to be the inertia among authors when it comes to OA. How do we persuade them to care and, as you suggest, to spend their money wisely when prestige and brand is the coin of the realm?
We found that most authors are okay with OA in principle, as long as they have the resources to pay the fees, can publish wherever they want, and see no change in quality as a result. This finding is really the backdrop for our entire project; it highlights the importance of bringing OA to the journals in which publishing is concentrated. Expecting authors to change where and how they publish as the primary route to OA in in our view a chimerical proposition.
As for spending wisely, we’re not at all suggesting that perceived journal value will cease to be a consideration for authors, simply that adding an economic dimension will influence their choices. Their price sensitivity suggests that they will spend their money wisely if it’s actually their money – if they’re involved in the cost/benefit calculation and are using funds that could be used for other things.
Your conclusions would seem to be at odds with European studies that are driving policy there and that seem to suggest a reduction in cost via Gold OA. How do you explain this?
The US is a very different market than Europe or most other parts of the world: institutions are mainly funded by tuition and philanthropy rather than government subsidies; more research funding is available in many disciplines; and negotiations with publishers are at the institutional or consortial level, not regional or national. So it makes sense that the economics of a different publishing business model would differ between the US and other countries. But we also note that other analyses are based on less data than ours, and make different assumptions about future APC costs. Finally, most of these studies have been focused at the national or global level. We knew that the economics would look different from an institutional vantage point, which is what matters for local decision-making.
The focus of the study is on journal articles, which are more common in scientific disciplines. Were there any clear disciplinary differences, especially between STM and HSS? And did you learn anything about the different challenges for long-form content and how we might tackle those?
Yes and no. Certainly author attitudes vary considerably across disciplines, with humanists far less accepting of APCs than scientists who already pay significant fees to publish (e.g. submission fees, page and color charges). And the availability of grant and other discretionary funds to authors varies greatly by discipline. But the actual APC costs don’t vary much today, perhaps because there are so few gold OA journals outside of STM, or perhaps because actual publishing costs are independent of discipline. This will be an important area for follow up research.
As for article length vs. long-form content, the same basic approach could apply, i.e., institutions provide authors with discretionary funds that can be used for publication fees among other things. The difference may be in the amount provided or the mechanism. For example, if an author in the humanities publishes two books with a publication fee of $15,000 each, they require access to at least $30,000. A scientist might publish twenty articles in that same time span, requiring an institutional subsidy of $1,500 each (assuming an APC of at least that amount), requiring that same $30,000 but perhaps on an annual basis. Since grant funding is far less common outside the STM disciplines, institutional subsidies for different publication types and disciplines will require careful thought.
Finally, I wonder if you can take a step back and tell us where this report and its conclusions fit into the wider landscape of gold, green and hybrid OA? What does it add to our knowledge of how to improve the openness of scientific and scholarly communication, and where do you think we can most productively focus efforts over the next few years?
The author survey turned up an interesting finding – in general, but particularly in the humanities, authors felt that if they couldn’t afford to pay APCs for the best journals, they would find alternative ways to publish their work. In the ‘publish or perish’ academic world we live in, creating incentives to change publishing practices is often discussed but rarely achieved. From our research, it appears that authors might be willing to pay high fees to publish their best papers in the best journals. But in other cases they might be more interested in finding alternatives – publishing fewer, better papers, or choosing low-cost options for the majority of their papers – and generally more conscious of what it costs to publish and what benefit they get from their choice. That might be a very good thing.
In terms of the landscape, green OA is clearly producing demonstrable value for readers today. A recent large-scale by 1Science and Science-Metrix documents the significantly greater citation impact that all forms of OA have over toll access, with green OA actually having the greatest impact at present. At UC, we see use of the articles in our institutional repository in all corners the globe. But green OA also relies on the subscription system for its lifeblood; at some point the two have to come together.
At the same time, initiatives like OA2020 and the recent EU open access declaration have precipitated a new conversation about prospects for a global conversion to gold OA – or what Danny Kingsley more neutrally terms ‘born OA.’ The growth of hybrid OA, which reflects the author preferences for existing journals that our study confirmed, represents both a challenge and an opportunity. Offsetting agreements being developed in Europe are an attempt to tame the hybrid phenomenon and turn it into a productive transitional strategy that leverages the efficiency of existing institutional relationships with publishers. Aligning the models of shared investment described in our study – via libraries, granting agencies, and author discretionary funds –with this approach, is an area that we’re very interested in exploring.
It’s also important to recognize that a conversion to born OA isn’t necessarily just about paying APCs to traditional publishers, but about freeing up subscription dollars to invest in OA initiatives of all sorts. While our study modeled a full conversion to APCs for analytical purposes, we see a much richer and more diversified open publishing ecosystem evolving in the future. Academic publishing behavior is still very conservative, but it may become less so over time. If we begin to utilize library collection budgets to subsidize author publication choices, our funds will be far freer to follow those evolving choices than they are today. That could be a major spur to innovation.