Woodstock crowd and stage
The crowd and stage at Woodstock, 1969. Image via Derek Redmond and Paul Campbell.

Usually discussion of open access (OA) has a utopian cast. It begins with the benefits to researchers and rapidly moves on from there–because it is a foundational premise that what is good for the research community is good for everybody. Expanded access will serve to cure horrible diseases, the science behind new technologies will cool a warming planet, and insights into people and power will make a veritable Woodstock out of the Middle East. Resistance, as the Borg say, is futile, but also immoral:  Who, after all, wants to stand between a parent and an afflicted child?

It is thus refreshing to see an examination of OA publishing that is sober and descriptive, one that examines how OA has been brought within the economy at large. This is what we have with Simba Information’s recent report, “Open Access Journal Publishing 2014-2017.” This is a clear-headed analysis of the business of OA, and it manages to survey the landscape without taking the idealists outside to shoot them. This is not the first Simba report I have had occasion to comment on here at the Kitchen. I was first attracted to them because I know the lead author, Eric Newman, who is a certifiably smart guy.  The report is a snapshot of OA (some of the story will be well known to Kitchen readers), its background, practices, and prospects. I will be curious to see if any OA advocates find anything in it that is wrong or unfairly presented.

Before we dip into the numbers, let’s acknowledge that research publishing, whether OA or of the traditional variety, captures value that is not always easy to count. While we can look at the public filings of the Public Library of Science and learn how much money is coming in, it’s harder to assess the economics of a journal that is hosted by an academic library, harder still to evaluate the economics of the parallel world of publishing that is Green OA. The full economy of OA, in other words, is still partly hidden. We can only count what we can see. I expect, though, that when all the numbers are ultimately revealed, we will find more hidden costs than hidden revenues. It also appears likely that what surplus or profit to be made from this form of publishing will fall to those who are already in the field or plan to enter it soon. OA, in other words, is beginning to mature and is thus subject to the same economic principles of any maturing business, including a tendency toward consolidation.

Simba notes that the primary form of monetization for OA journals is the article processing charge or APC. In 2013 these fees came to about $242.2 million out of a total STM journals market of $10.5 billion. I thought that latter figure was a bit high, and I’m never sure when people are quoting figures for STM alone or for all journals; but even so, if the number for the total market is high, it’s not far off.  That means that OA is approximately 2.3% of the total journals market (or is that just STM . . . ?).  That’s a big or small number depending on where you started from. With my personal bias in favor of new things, I see it as a huge number. Building a start-up is simply harder than running established companies. Bill Gates, Steve Jobs, Larry Page:  these fellows truly earned their money, but as for the chieftains sitting atop the large incumbents, it must be said that it is far easier to go from $1 billion to $2 billion than to go from zero to $10 million. The cumulative achievement of the OA publishers is an astonishing success story. Yes, you heard that on the Scholarly Kitchen, which some believe has a problem with OA publishing.

Simba sees OA rising to $440 million by 2017, which sounds reasonable. That would make OA 3.9% of the total market. That is neither a big number nor a negligible one. The more important point to make, which cannot be stressed enough, is that while traditional publishing continues to grow modestly, the OA portion of the market is growing much faster. Any publisher working in the research area would be remiss if they did not develop a strategy to tap into these growing sums. And of course just about all of them are.

The most interesting part of the Simba report to me is a table that lists the ten largest OA publishers by revenue. Springer, including its BioMed Central unit, is number one. PLOS is number two. By my count, two or three of the top ten are not-for-profit organizations. In other words, the revolutionary ardor that was supposed to come from the not-for-profit sector, where we fight for science for the sake of science, has run into the hard truth that commercial operators are good at what they do and know how to exploit an opportunity when they see one. And has there ever been a clearer opportunity than to rake in the APCs from funding agencies, which seem not to realize what it means to attach dollars to mandates that live outside the realm of end-user demand? The core proposition is that governments and funding organizations such as the Wellcome Trust are willing to pay for what librarians will not. You couldn’t make this up.

As I read through this report, with its many nuanced positions, including the implications of running out of research to publish, I was struck by the astonishing foresight and entrepreneurial genius of Vitek Tracz, who founded BioMed Central and got the whole ball rolling. Also dazzling in its performance is the not-for-profit PLOS, which not only navigated a tricky path through the marketplace, but also altered how we think about such things as the scope of a journal and the nature of peer review. I also began to wonder (I know I am not alone in this) if PLOS has pretty much played through its first inning and is going to have to pitch a different game going forward. There are several dozen organizations that are copying PLOS somewhat mechanically now, but will PLOS surprise everyone and launch an OA 2.0, which will enable them to sprint ahead of the prudent businesspeople who run the copycats?

I finished the Simba report just as Open Access Week drew to a close. The celebration had it right and had it wrong. OA has indeed established itself within the world of scholarly communications, though at 2.3% of the market, it is more of a toehold than a revolution. But the real story of OA is that as soon as someone found a way to make economic sense out of it, businesspeople jumped in and began to domesticate it for their own purposes. It is as though Thomas Paine were given a seat in the House of Lords. We will soon be looking back on Open Access Week as we do to Woodstock (I was there, by the way), with affection and nostalgia. As for where OA is headed, it’s just business.

 

 

Joseph Esposito

Joseph Esposito

Joe Esposito is a management consultant for the publishing and digital services industries. Joe focuses on organizational strategy and new business development. He is active in both the for-profit and not-for-profit areas.

Discussion

32 Thoughts on "The Size of the Open Access Market"

Before we dip into the numbers, let’s acknowledge that research publishing, whether OA or of the traditional variety, captures value that is not always easy to count. While we can look at the public filings of the Public Library of Science and learn how much money is coming in, it’s harder to assess the economics of a journal that is hosted by an academic library, harder still to evaluate the economics of the parallel world of publishing that is Green OA.

That is very true. But it omits what is surely the overwhelming part of the economic argument for OA: not the costs and revenues of the publishing businesses, but the indirect value of the published work on the wider world — as you say, “cure horrible diseases, cool a warming planet”.

Needless to say this value is not uppermost in the minds of publishers when they are running their business — they quite understandably have to think about the bottom line. But it surely is uppermost in the minds of goverments and granting charities when formulating their publication policies.

The real problem with this value is that it’s so darned hard to measure, or even to guess. Beyond my conviction that it’s enormous (and, no, I can’t give much in the way of objective justification for that conviction), what can be said? I truly don’t know, and would welcome pointers to any studies that have been done.

(I have more to say, but I’ll put it in a separate comment, since this one might hopefully spawn its own thread.)

Well, it is uppermost in the minds of mission-driven publishers like university presses and scholarly societies, Mike.

Yes indeed, Sandy; sorry for being inexact on this. My intended point here was not to assert that publishers are Bad Guys, but to acknowledge the existence and even legitimacy of perspectives other than my own.

Simba notes that the primary form of monetization for OA journals is the article processing charge or APC. In 2013 these fees came to about $242.2 million out of a total STM journals market of $10.5 billion. I thought that latter figure was a bit high, and I’m never sure when people are quoting figures for STM alone or for all journals; but even so, if the number for the total market is high, it’s not far off. That means that OA is approximately 2.3% of the total journals market.
[…] at 2.3% of the market, it is more of a toehold than a revolution.

Note though that this 2.3% is the proportion of the “market” in the sense of income. Many people are more interested in the proportion in terms of the number of papers published. Estimates of what it costs the community as a whole to publish behind a paywall and as OA vary, but I have offered estimates of $5333 for a paywalled paper and $453 for an open-access paper (the average of wildly varying APCs including zero). If those numbers are roughly right, then 2.3% of the scholarly publishing revenue equates to something like 22% of all published papers.

Note: I am well aware that the specific numbers are probably some way off the reality — these things are notoriously hard to estimate. But the broad picture, that OA by proportion of papers is something like an order of magnitude more significant than OA by proportion of revenue — is probably not far short of the truth.

(Note also, we are talking here only of Gold OA, ignoring Green.)

The cumulative achievement of the OA publishers is an astonishing success story. Yes, you heard that on the Scholarly Kitchen, which some believe has a problem with OA publishing.

🙂

The 2012 Outsell Report (http://goto.copyright.com/LP=123?source=copyrightdotcom) quoted in this recent post (http://scholarlykitchen.sspnet.org/2014/10/01/peak-subscription/) puts the number more like 10-12%, at least for 2012 (whether you think the percentage has doubled in one year is your call). That said, it does on the surface appear to offer significant value–2% of costs for 10% of the articles.

But it is unclear if this will scale (I say this to sound clever https://medium.com/conquering-corporate-america/10-tricks-to-appear-smart-during-meetings-27b489a39d1a), and whether this shows a representative 10% of the market. First, one must note that the articles here include a significant proportion of articles that are published at a cost that is higher than the revenue brought in from author payments (all non-PLOS ONE PLOS journals which charge $3000 per article but lose money, or eLife which spends $14K per article and just absorbs those costs). So the true costs are not shown by the revenue numbers.

Second, it’s unclear if the percentage here is weighted toward any particular market segment–maybe it’s capturing the part of the market that’s really inexpensive to produce, while being very light on the high end journals with higher editing costs and higher rejection rates. One can argue those things are unnecessary, but they are a significant part of the current market and can’t be ignored when trying to do direct comparisons, which are a bit apples to oranges here.

I think that 10% includes a lot of subsidized OA journals with no APC, making an APC per article figure misleading.

Why is that misleading? What’s at issue here is what it costs the academic community to publish articles and read them. A data point of $0 is surely just as legitimate a contribution to that as one of $3000 or $1350.

It’s misleading because it doesn’t represent the true costs of publication. If eLife spends $14K per article, then it’s misleading to suggest that eLife is magically free. There are real costs involved, and those have to be paid somehow.

There are real costs, yes; someone must pay them, in this case the eLife foundation. But in terms of analysis the cost to the scholarly community I think eLife’s price-point of $0 is just as valid as any other. (Otherwise you’d equally have to downgrade, say, PLOS ONE’s $1350 to some smaller number to account for its operating surplus.)

I suppose what this tells us is that when talking about money we need to be clear about the distinction between (A) the cost to academia, (B) the costs to the publisher — which are compensated more or less adequately by (A). In the case of eLife, the compensation is wholly inadequate (which they’re cool will because it’s their mission); in the case of the well-known publishers with 30-40% profit margins, the compensation is very much more than adequate.

Ignoring the costs of publication because it’s being subsidized doesn’t make those costs go away. Someone has to pay.

If the cost to the research community for a subscription article is (as you calculated above) $5K, and an eLife article means $14K in research funding no longer goes to the research community, then doesn’t that leave the community $9K behind? Even worse, much (most?) of that $5K doesn’t come from the research community, it comes from undergraduate tuition and student fees. By shifting the costs of publication from the library budget, which comes from a variety of sources at most institutions, and shifting it entirely to the research funding budget, that moves the majority of the financial burden onto the researchers, with them paying more than before. You’re also eliminating the contributions made by teaching institutions who do no research but use the research literature. Again, all the funds they contribute are now going to have to be covered by the researchers. So the overall cost to the academic community may stay flat, but the cost to the research community greatly increases, leaving less funds for doing research and more funds for important things like new uniforms for the football team.

And as PLOS is well aware, just covering the costs is not a sustainable model. There must be some surplus to allow for maintenance, upgrades, experimentation and improvements. If you just make exactly enough money to cover your costs, then some vital piece of infrastructure breaks, you don’t have any funds with which to replace it.

What major journals publisher is making 40% profits? or 35%? The highest I have seen is Elsevier at circa 30%. Thanks.

I’m hesitant to allow another comments thread on this to go too much further. These sorts of raw numbers are deeply misleading, and have been discussed in detail here:
http://scholarlykitchen.sspnet.org/2013/09/17/quantitative-comments-highlighting-some-of-the-math-about-finances-and-profts-contributed-by-readers/

Looking at a large and diverse company like Wiley’s overall profit margin and declaring the entirety of it to be from journals in flat out wrong, and as was noted in the linked post above, doesn’t follow corporate accounting practices. Worst of all, pointing out the earnings of the extreme outliers puts the rest of us, those at not-for-profits and university presses and research society publishers, in a bad light, as we don’t make (or desire) profits anywhere near the levels of some of the commercial houses.

Interesting to see that this report on Open Access costs rather a lot of money if you want to read the full text

I also began to wonder (I know I am not alone in this) if PLOS has pretty much played through its first inning and is going to have to pitch a different game going forward. There are several dozen organizations that are copying PLOS somewhat mechanically now, but will PLOS surprise everyone and launch an OA 2.0, which will enable them to sprint ahead of the prudent businesspeople who run the copycats?

For what it’s worth, I agree with this. PLOS has done a fantastic job, but it does rather give the impression of standing still now while the younger, nimbler likes of PeerJ and Ubiquity are making the moves. I’d love to see PLOS use some of that healthy operating profit to do something truly radical and exciting.

Open Access Week–just like Woodstock–hosted many different voices, from researchers, to publishers, to librarians, to funders. I wish we could have the same diversity of expression in The Scholarly Kitchen commenters. Unfortunately, a very small number of individuals with a very large amount of time on their hands have dominated the discussions around excellent posts like Joe’s (aka Country Joe McDonald). If we are to embrace the diversity the voices, perhaps it is necessary to impose limits to some of those individuals who won’t leave the stage.

Such is the nature of an open commenting system–to quote my son when rolling the dice during board games, “you get what you get and you don’t get upset.” We do moderate the comments here, cutting out the spam, anything inflammatory, and at the editor’s (read: my) discretion, anything annoying and unproductive. I do try to give people a free hand though, and if you’re going to target specific people and prevent them from commenting, why have comments enabled at all? It’s either a forum for discussion or it’s not.

Given that The Scholarly Kitchen is seen by some as cloistered and one-sided, I’d rather welcome in more voices than try to close the door on anyone. To me the choices are either you open things up and allow those who want to speak to speak (even if it means a regular crew of the same hecklers) or you just turn off the comments functionality altogether and let the posts stand for themselves (anyone who wants to comment can then do so somewhere else less connected and less visible). But I see little value (and don’t want to spend the time managing) a curated list of approved commenters.

I don’t know if the answer is to limit the participation of some–I’d rather see us figure out a way to increase the participation of others. I know there are many who have valuable and insightfult things to say, but who dislike the rough-and-tumble of our often “frank and open” exchanges and feel like life is too short to get embroiled in them.

For those reading from outside the Kitchen, this is something that we Chefs talk (and, yes, argue) about all the time. How do we retain the rigor and candor of this forum without driving away people who have good contributions to make? I’m genuinely of two minds on this. I think the rigor and candor are invaluable, and I wouldn’t want to water down our discussions in any way just to make them more comfortable–discomfort is sometimes the necessary price of dealing with tough issues. On the other hand, it really bothers me that I hear so many smart and insightful people saying “Comment in the Scholarly Kitchen? No way.” It seems to me like there must be a solution to this problem, and I wish I could figure out what it is.

Maybe this is a topic for a separate posting. I don’t want to deflect the commenting stream away fromthe content of Joe’s excellent piece.

Mostly a lurker, occasionally a commenter – I want to say I enjoy the way the Chefs and sous-chefs invest in this forum, and expand the content of posts through the comments. Where else would I get the chance to eaves-drop on such a fascinating group of people in this industry? It’s an acceptable price to pay – and an interesting part of the “social” aspect of blogging – that sometimes the comments swing onto a tangent.

Phil: your suggestion of ad-hominem censorship is beneath you, and beneath the Scholarly Kitchen. Whether you meant me or someone else, I hope you have had a chance to rethink, and that you will take the opportunity to retract your suggestion.

Meanwhile, if you disagree with my comments, you know the solution: reply, explaining why I am mistaken. As always, the response to free speech that you don’t like is more speech, not censorship.

It would be nice if Simba tried to estimate the OA book market as well. It is no doubt much smaller than the OA journal market at this point, but may well be growing at a faster rate just because it is newer.

Mike can you back up your convictions with some numbers. I spent some 40 years speaking with scientists in just about every discipline and none said they were reading articles on topology unless they were topologists and the same goes for various disciplines in biology, etc.. The great constraint on all these folks was time. It is hard for me to believe or have the conviction that more folks than those interested in a topic on a small sliver of science are reading OA articles on some topic in which they have no interest just because they are OA. On the other hand I can see a person reading a paper of interest regardless of it being OA or not.

In short, I do not see where OA has advanced the dissemination of information to those who most need it.

But, I do agree with Joe’s comment that OA is just business!

Harvey there is clear evidence that OA is an effective tool for increasing the dissemination of scientific papers. Take a look at this randomized controlled trial:

Davis PM. 2011. Open access, readership, citations: a randomized controlled trial of scientific journal publishing. The FASEB Journal 25: 2129-2134 http://dx.doi.org/10.1096/fj.11-183988

This clearly shows that OA has a major impact on readership of an article, with a doubling in html views, a 62% increase in pdf downloads and a more than 30% increase in unique visitors.

I beg to differ with the statement that OA has a major impact on READERship of an article, at least if one assumes that the readership consists of people who actually read the article Yes, more people download stuff when it’s free, but that doesn’t mean they read it, and even if they try to read it it doesn’t mean they understand it. For example, patients nowadays are coming into doctor’s offices with piles of documents about their condition downloaded from the internet, but my M.D. colleagues view this most often as a hindrance rather than helpful for the patient to understand their situation

Mike can you back up your convictions with some numbers.

If this alludes to “Beyond my conviction that [the indirect value of OA is] enormous (and, no, I can’t give much in the way of objective justification for that conviction), what can be said? I truly don’t know, and would welcome pointers to any studies that have been done.” — well, I made it pretty clear that I can’t put numbers on this. I hoped that other commenters would suggest ways to approach that measurement problem that I’ve not thought of.

But:

It is hard for me to believe or have the conviction that more folks than those interested in a topic on a small sliver of science are reading OA articles on some topic.

We have documented many instances of this on the site Who Needs Access?. Take a look and you will find nurses, teachers, artists, fossil preparators, small businesses, non-profit research organisations and many more who all need access to academic literature and find their work hindered by paywalls.

We can’t put a number of the opportunity cost of paywalling research, but we can see that it’s there!

OA has indeed plateaued. By contrast, electronic publishing went from circa < 5% market share of STM in the early 90's to well-over 50% by 2000 (and over 70% in law publishing). For OA to be at < 5% market share after 10 years of evangelical fervor is not indicative of a transformative much less dominant revolution; interesting, yes; symbolic, yes; important, probably; the way to major change over next ten years, likely not.
In other words, in the epublishing context the publishing industry and market broadly defined did in fact demonstrate the interest and the ability to implement a top-to-bottom upheaval. OA has fallen well-short of similarly being a massive game-changer (financially, 5% over 10 years is hardly statistically significant to an Elsevier or a Springer) even with the imperatives of the British Parliament, the California legislature, etc., behind it. Lastly, there is growing resentment among researchers at having to cough up precious grant funding to get published (see the Amer. Physiol. Asso, survey).
Predicting the future is a sure recipe for eating crow and I am prepared to do so in due course, but right now the collective plateau of OA is perhaps more like the now-diminished revolution predicted by MOOCs only a few short years ago.
(Oh yes: Vitek indeed has been quite the publisher for many years).

I think one very real contribution of Open Access is increased awareness and discussion of the publishing value chain. To paraphrase Paul Katzeff from Thanksgiving Coffee who speaks of… not just a cup but a just cup… maybe this is not just more business, but more just business. As for the longer term future it’s anyone’s guess, but questions are now being addressed that went unaddressed before.

It’s a very bad idea to politicize everything. It reduces a society to endless squabbling. The entire social justice argument concerning OA strikes me as silly. OA reaches researchers at major corporations and privileged academics at private universities. Let’s leave our pieties at home and focus on what something actually does.

political pious social justice squabble…I can’t respond in kind, but it does remind me of why the comments here are dominated by the same 5-6 people…

It’s a very bad idea to make everything about money. It reduces a society to endless squabbling. The entire economic argument concerning OA strikes me as silly. OA reaches researchers at SMEs, academics at minor, underfunded universities, and those with no affiliations at all. Let’s leave our finances at home and focus on what something actually does.

Comments are closed.