What if the Internet doesn’t disrupt scientific publishing, but merely leads to incremental changes in how it functions? What if there were a clear and strong practical link between achieving widespread open access (OA) and maintaining robust subscription businesses? What if a market containing both fresh young upstarts and strong incumbents thrived? And what if the large incumbents ultimately absorbed the upstarts?
What if the trade-offs many have been portending between big:small, old:new, and open:closed actually were dependencies?
A recent analysis of the scientific publishing marketplace focusing on the financial implications of OA policies and business practices presents these issues in between the lines, concluding that commercial publishers have weathered the storm and adapted to changes, making it unlikely OA would be much of a problem for them going forward, while suggesting that the ultimate solution to providing OA on a widespread and sustainable basis will depend upon a robust subscription market much like the one we have today.
The analysis, by HSBC’s Global Research division, is thorough and interesting, even if they do get some things wrong when it comes to market awareness. HSBC’s analysts surveyed funders, not publishers or librarians or authors, in making their assessment of the marketplace, which likely skewed the results somewhat and obscured some issues.
Their rationale for focusing on funders is stated toward the beginning:
. . . ultimately the Funders will foot the bill for Open Access and, although we regard it as improbably, the Funders also have the power to disintermediate the Publishers. In our view, this makes them the most important constituent in assessing the impact of Open Access on the publishers.
This assessment is purely a short-term financial assessment, not a long-term market assessment, which would likely focus on scientists and users of the literature, not on the funders. I’ll discuss the implications of this limited view below. The report also does not list which funders were surveyed, and provides no details about the survey instrument or methodology.
The analysis focuses on two publishers — Informa and Elsevier — while taking into account broader market trends, as well. Overall, the financial calculations make sense, and HSBC’s investment assessment is stated succinctly at the outset:
. . . journal profitability under OA should not be materially different to existing profitability, with a negligible probable impact on estimates.
Valuation anomaly has largely corrected: Although we continue to gain evidence that supports our view of a benign transition to Open Access, we note that the market’s concerns have also receded. From a short-term peak of risk aversion in January 2012 when researchers were petitioning to boycott publishers, the publishers’ shares have re-rated significantly. We find that valuations no longer reflect an “open access discount.”
Throughout the report, there are interesting summary lists and potent paragraphs. Here are a few of the highlights from the lists:
OA will not be introduced in a way that puts existing journal publishers out of business.
We found no appetite for regulation of APCs.
There is a risk that libraries lose more in funding than they save in subscriptions, intensifying the confrontation between libraries and publishers.
And here are a few highlight paragraphs:
“Open” access does not mean “free” access. The principle that taxpayers should not have to pay twice for research and that research should therefore be free at the point of use does not mean that publishers lose the ability to charge for their product.
Whether OA is achieved through Gold, or Green with embargoes, the publisher retains a revenue stream with negligible probable impact on estimates . . .
All of the funders in our survey agreed that APCs would be unregulated and set by the market.
Open access is hard for funders to enforce because they lack visibility on how many articles their grants are funding.
There are two points made I wish to highlight — the first, because I think the information is outdated; and the second, because I think their analysts are correct, and the implications of their statements are significant.
The first point comes from one of the lists in the report:
No funders foresaw that the role of publishers would or could be fulfilled by universities or funders.
The data in this report may have been gathered before the launch of eLife, a publication that has been created by funders. As noted above, there’s no way to know whether Wellcome Trust, Howard Hughes Medical Institute, or Max Planck were surveyed, but it would seem bizarre if they were not. If they were, and this statement reflects their responses, I can see only two reasons for this:
- Because eLife had not yet launched, they decided not to mention it.
- Because the funders believe eLife to be independent of them, they felt they themselves were not becoming publishers.
In either case, the truth isn’t as absolute as what the HSBC analysis states. As we’ve seen, eLife has been an extension of Wellcome, is competing as a publisher, and has prevailed upon another major funder (the NIH) for launch assistance. This is a wedge of disintermediation if I’ve ever seen one.
The second point I wanted to emphasize came from HSBC’s analysis of where OA is likely headed, given all the political, financial, and taxpayer pressures it faces:
In practice we believe that the majority of these mixed mandates are likely to be met through [embargo-based] Green rather than Gold, as Green is the path of least resistance and lowest cost for universities. . . . The length of embargoes proposed under Green OA ensures, in our view, that Open Access will be funded primarily by subscriptions.
This last point is one I’ve had on my mind for a while, ever since some conversations with a few OA publishers who acknowledge that, for selective OA journals, there is a fair amount of subsidization going on. In the larger publishing houses, the subsidies come from the subscription-based journals. In OA businesses, subsidies come from mega-journals like PLoS ONE, which do not have the same criteria as a traditional selective journal.
So, why might it be that “Open Access will be funded primarily by subscriptions”?
The first of these reasons stems from financial and political willingness, as in, How willing are funders or governments to be seen as supporting multi-billion-dollar publishing businesses directly? In this analysis, HSBC conducted an analysis of Elsevier’s business, hypothesizing that all of Elsevier might go to Gold OA. In their model, if Elsevier’s journals can command a 35% price premium on the market — that is, charge $4,050 instead of $3,000 for its APCs, on average — then it grows slightly. If it can charge 50% more — a $4,500 APC — then Elsevier would experience a 6.5% growth in revenues. If APCs are unregulated, and Gold OA is mandated, someone has to pay the bills. How will we get from here to there? And who will pay the bills? Right now, it would seem that Elsevier will do fine, and that APCs will be coming from funders and governments. Is that politically acceptable for anyone involved? Is it plausible in this marketplace? I think it’s difficult to imagine that Wellcome and the NIH would want to become known in the market as the primary funders of Elsevier, Springer, Nature Publishing Group, and Wolters-Kluwer, among others.
The second of these reasons comes from the cost:benefit proposition of Gold OA, which still seems weaker to me than advocates care to admit. That is, Are taxpayers willing to shoulder increased taxes to have immediate access to research papers? At least in America, we’re having trouble getting taxes raised on 1% of our population in order to pay for bridge repairs, food for poor children, health care coverage for the disadvantaged, and deficit reduction, I don’t see how something like this gets any floor time, much less gets knitted into the tax code. The benefits are too amorphous, and the trade-offs with research dollars too unpalatable.
The third reason is that I don’t believe anyone — funders, governments, or users of the literature — really want scientific publishers to be funded by the same people who funded the research. This has a “third rail” aura about it for most of the entities involved, and carries risks to reputation and legitimacy.
This report suggests a few interesting possibilities — that OA and subscriptions are linked in a very practical and long-term way; that big publishers have adapted to the incremental (not revolutionary) changes OA has introduced; and that the market and financial realities ultimately have a lot left to say about how things play out.
25 Thoughts on "Financial Realities — A New Analysis Suggests OA Will Have a Benign Effect on Publishers"
In your example above, why would you think Elsevier would have the ability to raise prices at will? My guess there would be more competition then you envision. Researchers currently can use grant funds for APCs but they do not have unlimited funds. Grants are for fixed amounts and PIs have some flexibility in how it is spent to achieve their objectives. If you spend and extra $1,500 to pay $4,500 instead of $3,000 you have that much less money to spend on other things. I think price will and does to some extent impact on the choice of APC funded journals and will help keep prices and profit margins reasonable.
Secondly. I’ve seen no one comment on the SCOAP3 experiment in physics Its a unique field but a very interesting model for a transition to OA. I am not sure it is easily transferable to other fields but an interesting way to shift the money libraries are currently finding subscriptions to providing OA and adequately funding publishers. I would be interested in what you and others think about the model.
Some APCs are already in the $5K range, and they are creeping up in many places. It’s a scenario, and not an implausible one.
Not every field will have the same path forward, another reason to accept a diversity of approaches. We’ve already seen that what might work for computer science or physics doesn’t work for humanities or biomedicine.
Elsevier might be able to charge $5000 for their premium label “Cell”, but certainly not for more than 90% of other journals …
I don’t know, I’d be willing to bet that, if the funds didn’t come out of the research budget, paying a large APC to publish in a top journal would be seen by many researchers as a good investment. Think of it this way–Nature has said that if they went pure Gold OA, then their APC would need to be around $30K per article. Given our current career advancement and funding structure, would publishing a paper in Nature result in an earnings and funding return of greater than $30K? In most cases, the answer is probably yes.
And as Kent pointed out below, Nature Communications already charges $5,000 per paper and sees a strong level of optional OA uptake.
the HSBC report and your analysis miss the point: in the current subscription world Elsevier and others are able to charge more than $5000 per article (mean) because they can sell bundles. But it is very unlikely they could keep the price in an OA world where they can only sell single articles. It is rather very likely that the price will dramatically decreasing for most of the articles …
The HSBC projections are based on current article publication levels. Elsevier only makes more money if they get to $4,500 per article. So, obviously, they’re not at that level now, nor at $5K.
Can you provide the actual numbers you used to reach the figures you’re stating?
According to Bernstein Research, Elsevier’s average revenues per article were $5,210 in 2011 (see p4, http://www.richardpoynder.co.uk/SCOAP3APCs.pdf).
There’s a huge assumption in the document you link to — that 50% of Elsevier’s revenues are from journals. If you take that down just 1% (to 49%) and use current exchange rates, it’s $4,819. So, it’s a very sensitive estimate, and 50% strikes me as a fairly arbitrary round number (as does 316,000 — again, adjust that upward by 1,000 articles, and you’re at $4,902, not $5,210. So, these dials jump a lot with just a couple of nudges. It may well be the same for the HSBC assumptions.
Thanks Richard, though that’s a very different number from what was mentioned above. The pdf you’ve linked to looks at overall Elsevier revenue, estimates that 50% of that revenue comes from journals, then divides that 50% of the total by the number of articles Elsevier publishes. That’s an extremely crude measure, but does estimate what Elsevier earns per article.
This is completely different from, what Elsevier charges per article. Journals are complex businesses, each with many different revenue streams. Journals earn money through subscription revenue, through advertising revenue, through secondary rights licensing (things like licensing articles to Ovid, EBSCO and the like), through commercial reprints, page charges, color charges, open access article processing charges, author reprints, translation rights, reprint rights, etc.
The only way to find out what Elsevier charges in the subscription market then is to remove all of the revenue sources above other than subscription revenue. For many journals, the non-subscription revenue is significant, some journals earning more from advertising sales than from subscriptions in particularly lucrative markets. Without that level of analysis, the $5,210 number tells you something very different. Even in a completely OA world, you’d still have many of those revenue streams in place and APC fees would not need to be at levels high enough to replace them all.
David – thanks, that’s well put. Earning different from charging. Lots of assumptions here of course.
By the way, the other oft-quoted source for the earning numbers is the Outsell industry estimates mentioned in the 2012 STM report, p5: “annual revenues from Eng-language STM publishing in 2011 was $9.4 bn, for c. 1.8 m articles”.
Now there is an interesting triangulation you can make with some other publicly-quoted Outsell estimates in the Economist (‘Changing Nature’), which suggest that Outsell think journal *subscriptions* gleaned $6 billion for the industry in 2011.
On that basis, 64% of STM publishing revenues are direct from subscriptions, so that Elsevier might be expected to be making $3,333 in per-article charges and the rest from other revenue streams. But I think Outsell are not including society membership fees, which in effect can be subscriptions. It’s all too rough-and-ready to mean much, but highlights the difference between total revenues and what APC revenues may have to be.
I think there are two basic problems, insufficient information and sloppy analysis. It’s difficult, if not downright impossible to parse financial reports from both private and public companies to get the details one really needs to do accurate modeling. Companies tend to obscure this sort of information. I’m not really sure if there’s a specific goal in doing so, perhaps just general business paranoia, not wanting to let the other guy know exactly what you’re doing as it may give them some sort of unknown advantage. But more transparency on the part of publishers would certainly help everyone make rational, data-based proposals.
And the other side of the coin continues to amaze me. Reputable scientists, who religiously follow rigorous experimental design and scientific methods in their own research, seem to throw all of that out of the window when they make arguments about or proposals for publishing. They’re willing to omit controls and make assumptions about unknowns, things they would never do in their scientific work. The approach is “I’ll see it when I believe it,” rather than the other way around.
As part of the maturation of the OA movement, we need a higher level of rigor (and the accompanying better data mentioned above). As we move from the theoretical to real world implementation. Vaguely fudging the facts is not good enough.
You touched on something that I have been struggling with, that third rail issue. It bothers me to think of a future where a funding body could decide whether they will pay for an article to be published. There is an understanding now that you get the grant, you do the work, you report the results via journal article. If the funding body decides it does not like the results, what happens then?
Exactly. We’ve seen government suppress scientific reports that did not back the administration’s position. What if that were to happen in journal publishing?
Don’t forget also that many funding bodies are planning to offer OA fee money in the form of block grants to universities. That means adding in a new level of pre-publication peer review, where a university committee will review every researcher’s manuscript and decide whether it’s worthy of publication and whether they have chosen an appropriate journal for the submission.
Indeed, and this has huge implications for intra-university politics. E.g., if a junior and senior faculty member both have papers to publish in a journal and there is only enough money in the pot to pay for one APC, who do you suppose will be awarded the money?
As the headline of this post notes, OA may end up being benign for publishers . . . but, by inference, it may not be benign for governments, researchers, academic institutions, or funders. Issues already appearing include academic freedom, allocation of funds, qualifications of disbursement agents, delays caused by up-front funding, and conflicts of interest.
Kent, you don’t address the other prong of the last claim, viz., that funders don’t think universities will be able to take over the role that publishers perform. But this ignores over a century of direct engagement by universities in scholarly publishing through university presses, which began with Johns Hopkins publishing STM journals, and it also ignores the burgeoning interest of libraries in becoming publishers, as with the new 50-member Library Publishing Coalition supported by the Educopia Institute: http://www.educopia.org/programs/lpc.
You’re right. As I noted, there was a certain amount of marketplace ignorance in the report, which is to be expected given its approach. Funders aren’t proving to be very savvy about publishing, I’d observe. Even in the digital age, initiatives like HighWire Press and Project MUSE are standouts, and there are plenty of other things going on (repositories, arXiv, etc.). Ultimately, though, these seem more incremental, not revolutionary, which I think is the overall point.
IMO as the infrastructure needed to publish OA grows the APC will go up. At some point funding agencies will question the goals of OA, authors will question the amount paid to provide an OA article, the public will question if it is worth the cost to have the information available to a few for such a steep price, and we may see a newly launched model called subscription based publication.
On another note. Kent do you see the trust creating a publishing company to handle ELife much like ACS has a publishing division?
David, your observation is a good one in that on a cost benefit analysis to the author the fee for publishing in Nature if is should charge one is beneficial.
I don’t view Wellcome creating eLife as equivalent to ACS having a publishing division, not at all. Societies and publications emerged to answer a need professionals had around sharing information and findings. Publications were the asynchronous part of societies, and still are in many cases. Meetings and colloquia were the synchronous parts, and still are. Wellcome is a funder of research, not a society of professionals or scientists. It has a very different role, and one that I believe conflicts with the roles of publishers and editors.
I do agree the virtues of the subscription model may well reassert themselves pretty soon. For many years, all this fuss has struck me as what my Dad used to call, “learning the hard way.”
I agree with your analysis of Welcome competing in the marketplace. But, say they formally form a publishing house (which they are now). That they split it from the formal trust. Wouldn’t this be a way of legitimatizing what they are doing?
What amazes me is that the AAP has not weighed in on their activity.
I’ve said as much. The Gordon and Betty Moore foundation gave PLoS a few million dollars to start up, then walked away (for all intents and purposes). That created a firewall. Wellcome and HHMI and Max Planck seem to want to keep their hands in for some reason, and each editor was apparently chosen by one funder, and each of the three has a lot of funding from their funder. It’s very conflicted.
I think the silence in the community stems from the power the funders have. People may be afraid to speak up. I know there’s a lot of background grumbling and concern, however.