Spanner in the Works
Spanner in the Works (Photo credit: Kapungo)

Scholarly journals rely on a complex mix of revenue streams. The complexity of the system makes it difficult to directly model the economic impact of major upheavals, such as the current trend toward public access mandates from research funders. Revenue sources other than subscription fees seem poorly understood and rarely acknowledged by those setting policy, but they need to be factored in to policy decisions and publisher financial planning.

In yesterday’s post, David Wojick presented an approach for analyzing the impact of funding agency public access policies on scholarly journals. As noted in the comments, I feel this approach is flawed for a variety of reasons, one of which is that it doesn’t deal with revenue streams beyond subscriptions. This does not accurately reflect the way that many (most?) journals earn their keep. Though it is difficult to tie down specific numbers, estimates suggest that somewhere between one-quarter to one-third of STM journal revenues come from sources other than academic subscriptions.

That’s a pretty large amount of revenue that needs to be accounted for, and it’s a particularly important set of revenue because much of it comes from outside of the research community.  As we seek favorable new policies, one goal is to keep research funds focused where they should be–paying for research. Any support for journal expenses that shifts the financial burden off of the researchers is very welcome.

Much of this non-subscription revenue, in my experience, comes from what’s often called “secondary rights” licensing, primarily licensing the distribution rights for published papers to aggregators like Ovid, EBSCO, ProQuest and JSTOR. I’ve worked with journals where aggregator rights make up less than 1% of total revenue, and journals where they comprise more than 50% of what the journal brings in. A smaller, but still significant stream is pay-per-view (PPV) revenue from access to individual articles.

Both these revenue streams would seem to be directly threatened by public access policies mandating free availability of articles after an embargo period.

When the only articles that can bring in PPV funds are those less than 12 months old, that drastically cuts inventory and sales rates – especially if it is widely known that the content becomes free in a matter of time. What happens to an innovator like DeepDyve, which offers article rentals for a low rate to non-subscribers, when their rental inventory will be massively reduced?

Journal aggregators provide access to a broad array of material, including journal articles from many publishers. The journal articles are not immediately available upon publication, but enter these programs after an embargo period, typically one or two years. The services offer lower cost access to an enormous number of journals to institutions that want to offer these resources to students and others, but that don’t need immediate access to the newest and latest articles. These are often schools where research programs are not the primary focus, or libraries seeking to broaden their holdings by including subjects outside of the core research areas of the institution.

If most articles are made freely available, either immediately or after 12 months, then there seems little reason to continue to pay for the journal content in these sorts of packages. That may spell doom for this segment of the aggregators’ business models and an end to an important source of financial support for journals. It’s perhaps not surprising to see aggregators diversifying and investing heavily in new ventures, as these packages represent an important tentpole and their loss will be a significant blow.

Consider also lost revenues from the Copyright Clearance Center where policies requiring a Creative Commons license would allow unpaid reuse of articles. Suddenly, the revenues from this source also dry up. Together, these various streams all add up to a significant portion of journal financial support.

Some will clearly benefit from these policies. The teaching institution that can drop their aggregator subscriptions can instead put those funds to other needs. The pharmaceutical company that wants to use a figure from your paper in their advertisement no longer needs to pay to do so. But those savings have to be offset somewhere, and that likely means a larger financial burden shifting to the research community.

When a business loses one revenue stream, it increases the demands on the others. As policies continue to limit the ways that journals can cover expenses, more and more pressure is placed on the few revenue generating activities allowed: subscriptions and open access article processing charges.

We already see publishers charging increased fees for CC-BY licenses to replace lost secondary rights revenue. It seems counterintuitive on the surface, but policies that broaden public access to research articles may indeed result in higher subscription prices. Measures taken to alleviate the “serials crisis” may instead make it worse by limiting financial models for journal support to channels that come directly from researchers and their institutions rather than outside sources.

These secondary effects of public access mandates don’t seem to be accounted for in most policies. Embargoes remain an important tool to help preserve subscription revenues, but they may not be long enough to preserve secondary rights revenue. Complex systems are difficult to model, particularly in conditions of major upheaval, and most policy seems based on a fairly rudimentary approach to journal finances.

A reduction in secondary rights revenue seems almost certain, and research societies and publishers need to factor this in to long term financial planning.

Enhanced by Zemanta
David Crotty

David Crotty

David Crotty is a Senior Consultant at Clarke & Esposito, a boutique management consulting firm focused on strategic issues related to professional and academic publishing and information services. Previously, David was the Editorial Director, Journals Policy for Oxford University Press. He oversaw journal policy across OUP’s journals program, drove technological innovation, and served as an information officer. David acquired and managed a suite of research society-owned journals with OUP, and before that was the Executive Editor for Cold Spring Harbor Laboratory Press, where he created and edited new science books and journals, along with serving as a journal Editor-in-Chief. He has served on the Board of Directors for the STM Association, the Society for Scholarly Publishing and CHOR, Inc., as well as The AAP-PSP Executive Council. David received his PhD in Genetics from Columbia University and did developmental neuroscience research at Caltech before moving from the bench to publishing.


28 Thoughts on "Indirect Economic Impacts of Public Access Policies for Journals"

Very useful, David. However, contrary to what you say, there is nothing in my approach to modeling the adverse economic impact of embargo based mandates that precludes including the losses you describe. I merely started the discussion with subscription losses, but we have yet to build the model. If your numbers are correct then the loss of secondary revenues should be a major component of an adverse impact model for the subscription industry as a whole, but not necessarily for individual publishers.

Regarding making up for secondary loses by raising subscription prices, there is no guarantee that this will work. This is because raising prices should reduce sales, especially if these price increases vary greatly from one publisher to another, as your numbers suggest they would. If you lose half of your income, you have to double your price. The real solution is to have reasonable, evidence based embargo periods. Phil Davis’s numbers suggest that twelve months is far too short.

Acceptance of increased subscription prices will vary journal to journal. If a journal is peripheral to an institution’s focus, it may be dropped. If it’s a vital, key journal, then you may see the increase accepted and paid for by dropping an unrelated journal that’s not seen as important by faculty. Again, something indirect and hard to model.

I’m also not sure how embargoes would help here. We’re talking about content that is already several years old.

Economic models are statistical in nature so the fact that the distribution is wide need not make it hard to model. If longer embargo periods do not change the secondary revenue impact then that actually simplifies the model by reducing the variables, making secondary impact a constant, as it were.

This is because raising prices should reduce sales, especially if these price increases vary greatly from one publisher to another

In a normally competitive marketplace this would be true, but scholarly publishing is not a normal marketplace. Raising subscription prices only reduces sales where a substitutable alternative exists. If Toyota raises the price on all of its cars by 50%, you can buy a Ford or a Nissan instead and you get a product that is, in all important respects, functionally the same. But when Nature or the American Chemical Society raises its subscription prices by 50%, you don’t have the option of getting virtually the same content from anyone else. This makes demand for core-content subscriptions extremely inelastic: you can’t substitute cheap RSC content for expensive ACS content, because it’s not the same information (even if it’s in the same discipline).

It is indeed a matter of elasticity but surely the number of journals that are infinitely valuable to all subscribers is quite limited. I just dropped my subscription to science because I was not reading it. If they doubled their price I am sure others would do likewise. And those who do not might drop something else, perhaps from the same publisher, so the losses are still there.

It’s true that there is a finite number of journals for which demand is extremely inelastic. But while that number is finite (and varies from library to library), it is nevertheless fairly large, and the publishers of those journals tend to raise their prices aggressively from year to year. What this leads to is not cancellation in favor of cheaper substitutes (which don’t exist), but the cancellation of less-core titles in other disciplines. Since journal publishers tend to focus on a limited range of disciplines, the risk of loss from such cancellations is pretty minimal in most cases — and even where it exists, not all cancellations are equally damaging to the publisher. A publisher with a wide range of titles may jack up the price of its expensive physics journals by 40% and still come out on top when it loses some of its cheaper humanities subscriptions.

Your experience with Science reflects the fact that Science is different from most scholarly journals. Your interaction with Science takes place in the marketplace of consumer magazines like Time and The Economist. While they too offer technically unique content, their content is more readily substitutable–an article in The Financial Times on global currency trends may be roughly substitutable for an article on the same topic in The Economist. In the realm of scholarly journals, such substitution tends not to work, because multiple articles aren’t available from different journals reporting results of the same experiment or study.

You’d also have to factor in that many (most?) of the journals you’re talking about are going to be part of collections and journal packages, so attaching specific numbers to specific titles becomes more nebulous.

Right — journal packaging effectively binds the fate of non-core titles to that of core titles, making cancellation in response to price increases even more difficult (and less rationally applied). This really frustrates faculty at times. They wonder, with good reason, why the library canceled a core title in their discipline but is still subscribing to Cupcake Decorating Letters B.

Once again, economic modeling is statistical in nature so one does not attach specific numbers to specific journals. The exception would be a relatively small publisher that wanted to explore their potential for adverse impact, but there one could probably make reasonable estimates on a per journal basis.

I specifically listed bundling in my Kitchen article. My thinking on the embargo versus subscription side is that Phil’s results suggest that most disciplines have long half lives so bundling is no protection. I had not thought about large price increases due to the loss of secondary markets, but I am not prepared to believe that journals are somehow exempt from economic decision making. If available funding is constant and prices are increased dramatically then something has to give. The modeling question is what will it be? If this is complex, and especially nonlinear, then that is all the more reason to try to model it, not less, because intuition alone cannot handle it..

I am not prepared to believe that journals are somehow exempt from economic decision making.

No one is saying that they are. But price-and-demand dynamics are not the same in every marketplace, and the scholarly publishing marketplace is a uniquely weird one.

If you read my comments to your prior article you will see that I did not agree then and I do not agree now. I have done regulatory impact analysis in a lot of different industries and each thinks it is somehow uniquely above analysis, but it is not so. The journal industry is actually fairly straightforward, compared to some. For example, the Navy once tried to develop a unified accounting system for its laboratories. They spent nine years and half a billion dollars, then had to scrap it because it did not work. Military R&D funding is a truly weird system. But we can agree to disagree.

David W: I am not sure that authors and journals are not protected from using materials in course packs or readers. I think one can read the material but I am not sure they can re-sell it.

Perhaps I am wrong.

I have done regulatory impact analysis in a lot of different industries and each thinks it is somehow uniquely above analysis, but it is not so.

To my knowledge, no one is suggesting that the scholarly-journal industry is above analysis.

For some publishers this is certainly the case and all of this needs to be part of the discussion. At ASCE, for whom David Wojick’s report was done, our PPV income is not a significant source of revenue. That may be changing, however, as we try to convert lost individual member subscriptions to our inexpensive download bundle packages. The download cards are diverse as the available content also includes conference papers and ebook chapters. Those two content revenue sources seem safe for now.

Aggregator fees is an interesting topic. We have been careful not to allow anyone to sell full text access to our content. We tried allowing content over a year old to be sold via an aggregator’s collection and we lost subscriptions to the journals. The loss in subscription revenue, and the loss of having direct access to our customers, did not equal the royalty amount. Remember that Civil Engineering content usage is heavy on content older than a year. I would agree that for some (many?) journals that do make significant money on these deals, losing that income with little opportunities to recoup that somewhere should be part of this conversation.

What this conversation yesterday and likely today really shows is that a one size fits all model is not at all appropriate. And we can’t just slice it in two categories: STM vs. Humanities/Social Sciences. Within these groups, there are many flavors of business models and realities.

Discussion is the operative word at this point. Sometimes the real payoff in trying to model a complex process is that it gets one to sort out and think about each of the factors involved. You may recall that in my original analysis of the US public access program I pointed out that having different embargo periods for different disciplines was perhaps the greatest challenge, one the Feds would shy away from. At least we are now addressing that challenge analytically.

Institutions of Higher Education (IHE), as they play their dual roles of primary source of this kind of content and primary consumer of this kind of content are also under pressure. Inelastic demand (include textbooks as well) has pushed increases in the cost of tuition beyond those of even health care. IHEs are under great and growing pressure to correct that trend, especially publicly supported institutions. Legislatures have slashed appropriations to higher education.
Yes, publishing has certain requisites that really do cost money and, yes, someone has to foot the bill for it. However, allowing the intermediation of cost-inefficient publishers may no longer be sustainable or tolerable. IHEs could dis-intermediate scholarly publishing. This has already begun with eTextbooks and the lessons learned there may prove equally useful in driving down the costs of journal subscriptions, author fees, etc.

I find it hard to believe that the library budget is causing an increase in the cost of tuition. Perhaps the author above can support the claim with some data.

Reply: The CPI from 1980-2012 shows tuition increased faster than any other category, including health care (chart: description:
According to the ALA (2005) (, the subscription prices of scholarly journals have been increasing at a rate faster than the CPI-U which was 179% between 1980 and 2012. A more recent survey (2012) by Allen Press: tells essentially the same story.
Historically, the annual increase has been 7.3% so to compare the rise in journal subscription prices with the rise in tuition from 1980-2012, we get 234% which is far ahead of the 179% CPI-U (consumer price index for all urban consumers)
I did not say that library budgets are the sole cause of tuition increases. For public IHEs, reduction of state appropriations (tuition subsidies in reality) have been brutal (see:
Reducing the tuition and fees (including textbooks) that translate into student debt is a very high priority among public IHEs. They are already chipping away at textbook costs and the same technologies and tactics that are enabling that effort can and probably will be used to help reduce the pressure of journal subscription costs on tuition.

There is no doubt that subscription prices have increased but there is no data that shows that tuition increase is tied to subscriptions pricing. Thus, I would still like to see data that supports your contention that tuition increases is tied to subscription pricing.

Why are you conflating textbook costs with subscription costs with costs of higher ed? You are making great leaps of faith. You have failed to mention the cost of beer, which is a substantial cost of IHE!

Some supplemental revenue already disappeared in the transition from print journal publishing to electronic. E.g., university presses that joined Project Muse (and for some this was the ONLY way they could afford to go electronic) saw much of their advertising revenue disappear, as the electronic version became primary while print subscriptions dropped and Project Muse did not include advertising in its e-versions of the journals it aggregated. The loss of revenues from reuses of articles in coursepacks, e-reserves, print anthologies, etc. with enbargoes extending back only a year could be significant for journals in the humanities and social sciences. E.g., one of the strongest revenue-producing articles for the journal Philosophy & Rhetoric appeared in the very first issue published decades ago, but it was reused many, many times and the author made many thousands of dollars from the fees shared with the journal 50/50. Loss of that kind of revenue would be a loss not only to the journals but also to some authors like this one.

David, I looked at your data link, which is to a two paragraph blog comment — It does not say anything about secondary rights revenues. The second paragraph has a detailed breakdown with subscription revenues at 93% plus advertising and APC type revenue, with no mention of secondary sales. The first paragraph seems to contradict this breakdown somewhat, with subscription revenues of just roughy two thirds of total revenue, but no breakdown of the other third. Again there is no mention of secondary sales.

The contradiction illustrates nicely some of the data problems that need to be dealt with, but they do not seem to support your suggestion that the loss of secondary rights sales looms large in the adverse economic impact of embargo based mandates. What are we missing?

Data availability is indeed a major problem in any analysis of this sort. You have a large number of companies with very different accounting practices. More importantly, they are all in a competitive market and few, if any, are going to release any more data than they are required to by law.

The data you’re seeking is seen as confidential. It was something of an impressive feat that Phil was able to get usage data, even anonymized, out of some of these companies. Financial data? Good luck with that.

My understanding, both from the cited sources and asking colleagues, is that the quoted numbers are fairly accurate. Somewhere between 20-40% of journal revenues are coming from sources other than subscription.

I do have access to my company’s journal finances but I’m not at liberty to publicly release that data. As noted in the post, I can attest to the great variability that different sources play in funding different journals. Some rely heavily on secondary rights, some on advertising, others none of the above.

But the quoted STM report apparently says that just 7% of revenues comes from non-subscription sources, not 20-40%. Are they wrong?

I haven’t seen their source material, so can’t tell you. All I can tell you is what I know from the material I have access to, and the numbers I’m familiar with are higher than that. Perhaps another level of complexity on the subscription side though–what are the economic pressures determining subscription decisions in a corporate setting or a government setting and how do those differ from an academic setting?

You might want to explore getting these data from university presses attached to public universities whose financials are presumably open to the public as required by law.

Good point Sandy, but at this point we may not need a lot more data. When an agency does an RIA they typically spend several million dollars, including a lot of data collection and industry analysis. At this point what the subscription industry needs is a prima facie case that an RIA is warranted because there is the potential for significant adverse economic impact from the US public access program. Rough data may be good enough for that.

Comments are closed.