Tuesday, June 25, 2013

The Big Deal

Andrew Odlyzko has a fascinating paper with a rather long title, Open Access, library and publisher competition, and the evolution of general commerce (PDF). He describes how the relationship between the libraries and the publishers in the market for academic journals has evolved to transfer resources from libraries to the publishers, and how a similar strategy might play out in many, more general markets. Below the fold I discuss some of the details, but you should read the whole thing.

Andrew starts by using statistics from the ARL to show that, between about 2003 and 2010, the number of serials available in research libraries increased about 4-fold as a result of libraries switching from individual subscriptions to "Big Deals" granting access to the whole of a publisher's catalog.
ARL statistics for the academic years 1989-1990 through 2009-2010 ... show that the last decade has produced a remarkable increase in availability of serials. ... The increase in numbers of available journals was also accompanied by a notable leveling in availability among ARL members. ... There is a great degree of compression, especially at the top, with the 90-th percentile institutions far closer today to the median than in earlier years. The distribution of library budgets has not changed much over this period, so what we observe is the result of increased price discrimination. ... the “small” institutions gained more than the “large” ones. ... The average price per serial received by ARL academic members has plunged over the last decade, to a level last seen around 1990, even in current dollars, without any adjustment for inflation.
The combination of more journals at lower prices resulted in slow growth of the total spending by libraries:
The growth of the last decade was certainly made possible by improvements in technology, so that huge collections of journals can be made available in electronic versions even by libraries that could not possibly hope to house them physically. But much of the credit must be assigned to the reviled “Big Deals,” in which publishers enriched their offerings by tossing in many additional journals. This increase in journal availability was not accompa- nied by commensurate revenue increases. As is shown in the ARL statistics summarized in Section 5, while the growth in spending on serials did outpace the growth in library budgets, it did so by a smaller margin in the first decade of the 21st century than in the last of the 20th.
These deals look different from the perspectives of the three parties involved, researchers, publishers and libraries.

Researchers generally view the "Big Deals" as a positive development; they get access to 4 times as many journals. Few are aware of the increased cost, which in most cases is confidential. Even if they are, it does not come out of their budget. Even fewer are aware that the additional journals obtained for the extra cost add, on average, little value. These journals have been created by the publishers to inflate the size of their Big Deal. More journals in a "Big Deal" increases the value librarians perceive because they believe that the number of journals is a measure of value. It also provides the publishers additional editorial board slots with which to bribe researchers. To oversimplify, the same amount of research is being spread over more articles in more journals, a point that Andrew seems to miss.

Publishers, at least the large commercial ones, think the "Big Deal" is a brilliant idea for several reasons:
  • It places their competitors, the individual society publishers, at a serious competitive disadvantage. They cannot themselves offer a "Big Deal", lacking enough titles, and their small contracts covering few journals will be first in line for cancellation as the price of the "Big Deals" increases. Recent years have seen many society journals being sold to big publishers.
  • It allows for massive price discrimination, selling the same product to different customers at different prices, which are in practice the most the customer can possibly afford. Andrew uses the example of Elsevier's deals, in which Harvard buys the same product as the University of Montana but pays nearly 5 times as much. Price discrimination by monopolists is a well-known way to extract rents.
  • It is a bundling tactic, another well-known way for monopolists such as Microsoft to extract rents by forcing customers to pay for unneeded products, and blocking competitors. In this case these products are the numerous low-quality journals created to pad out the bundle.
  • It drives libraries to cancel print, transforming their relationship to the publisher from owning a copy to renting access. Libraries renting access to the literature from a small number of publishers are in a very weak negotiating position when the rental contract is due for renewal.
  • It suppresses the demand for Open Access. Researchers at libraries with "Big Deals" have unfettered access to the vast majority of the literature. Even universal Open Access would make little difference to their lives, so what is the point of campaigning for it?
Libraries are on the horns of a dilemma. The "Big Deals" are an easy way to satisfy the demands of their researchers for access to the literature. But the "Big Deals" eliminate almost all functions of the library, except negotiating the "Big Deals" with the publishers. Any individual library is a small customer for a big publisher, and the "Big Deal" removes their ability to walk away from the table, thus they have little if any leverage. "Big Deal" negotiation is much more effective at a national level, as in the UK. Andrew demonstrates that the result of the switch to "Big Deals" has been to increase the proportion of the library budget going to the big publishers, and reduce the proportion going to traditional library services. Universal Open Access would remove much of the funds passing through libraries' budgets and their last major function, so it might be even worse for libraries than the "Big Deal".

Andrew is relatively sanguine about this effect on library spending, perhaps because he over-estimates the value of the additional journals:
In the short (and even intermediate) run, “Big Deals” therefore do promote wider access to scholarly literature, and they do serve to push libraries to be more efficient. Thus what we observe in the ARL libraries is Adam Smith’s “invisible hand” producing socially desirable outcomes. In the long run, of course, “Big Deals” do entrench the publishers, their profits, and their inefficiency.
Andrew points out that the decline in traditional services is probably an inevitable result of technological change; it is clear that libraries are failing to compete with generic Web services such as Google:
There are many interesting statistics ... demonstrating decline of the traditional functions of libraries. Thus between 1995 and 2010, the number of students at ARL institutions grew by 33% (with the ranks of teaching faculty and graduate students climbing 15% and 43%, respectively). The only category of library services involving physical material that showed growth was interlibrary loans, which climbed 92%. This reflects libraries concentrating their budges on serials, and giving up on trying to keep up with the growth in the number of new books being published. In other categories, initial circulation ... of physical volumes dropped by 42%. Thus it is a gross exaggeration that “nobody uses the library anymore,” as one sometimes heard from faculty or students. But the decline in borrowings per student by more than half is telling. What is perhaps most surprising is that the number of requests for reference assistance dropped by 66% in absolute terms, as is shown in Fig. 6, and thus by about 75% on a per-student basis. This is certainly a core competency of librarians, and they are great at navigating the torrents of electronic information, as well as providing guidance to the use of traditional printed sources. However, it appears that Google, Wikipedia, publisher databases, and the like are “good enough” for most scholars, and that the convenience of around the clock access from anyplace outweighs the higher quality that librarians provide.
Andrew is optimistic about future opportunities for the big publishers:
The basic and very promising approach open to publishers is to continue marginalizing libraries by extending the reach and scope of “Big Deals.” The consortium model, in which groups of libraries cooperate to get access to a “Big Deal” is already common, and can be pushed further. The ultimate situation might be national “Big Deals,” where some top- level bodies pay for access for everyone from a nation. Enlarging the “Big Deal,” especially through further mergers, but also by including additional information sources, can serve to create packages that simply could not be dispensed with. The most obvious move in that direction (which is already taking place to a small extent) is to make books, both current and old ones, a part of the “Big Deal.”
While I agree with Andrew's bleak view of the future for libraries, I think his view of the future for publishers is too rosy, for two reasons. First, in the Web world, journals have lost their function as routes for access to the articles. Readers access articles directly via search, links and RSS feeds rather than via the issue table of contents. The remaining role of journals is as tags that publishers attach to articles in their vast databases of articles. These tags are supposed to indicate the quality of the articles but they do not. It has been conclusively shown that the quality of an article is not predicted by the journal that publishes it. Indeed, the peer reviewed research on peer review shows that, once the value that highly selective journals subtract by encouraging fraud and misrepresentation is accounted for, it is unlikely that any journals add detectable value. This is driving the "altmetrics" movement, aimed at measuring the quality of articles directly. Altmetrics are a problem for the publishers. The set of journals included is the only way publishers can distinguish their "Big Deal" from their competitors' "Big Deals". The more the world judges article quality directly, the more generic the big publisher's products, and the less justifiable their massive profits. After all, were journals to be recognized as meaningless tags, the publisher's only remaining function would be to run their databases of articles. Why pay vast amounts to have 5 separate databases each with 1/5 of the world's articles?

The other problem for publishers is that they might actually succeed in making libraries obsolete, and thus have to choose between selling to individual researchers and selling to governments. Selling their content directly to researchers would eliminate price discrimination, and would drive universal Open Access almost overnight. Governments would be a negotiating partner with much greater leverage than individual libraries. Unlike individual universities, governments provide most of the funds that pay for research, so they are in a position to enforce Open Access mandates effectively, and take other measures, such as making it clear that faculty authored articles are "works for hire", that would destroy the publisher's business. On the other hand, the small number of large publishers with vast profits are in a strong position to "lobby" politicians to continue the flow of public and tax-deductible money their way.

Of course, many researchers are broadly aware that the large publishers are making extraordinary profits from their work, and some are protesting. In some fields, primarily those using the arxiv.org pre-print service as the primary mode of communication, the reduction of journals to mere tags supposed to denote quality is evident. In these fields some publishers, for example Usenix, have gone open access and regard their publications as an adjunct to their real business, running workshops and conferences. This business is much less vulnerable to technological dis-intermediation than journal publishing.

Andrew's most thought-provoking point is that academic publishing may represent the future of the general economy:
A frequent complaint by researchers and librarians is that publishers profit from the unpaid labor of scholars who write papers and referee submissions. Yet that is an increasingly common feature of the economy, and scholarly publishing may be viewed as a forerunner of this trend.
He points out that Apple's iPhone and iPad sales are driven by the multitude of apps available on the platform. The app development market is long-tailed; a very few apps make a lot of money but the overwhelming majority of app developers make very little, so Apple is dependent upon the effectively unpaid labor of its users:
The Apple app developer situation is characteristic of many of the most visible recent high-tech success stories. Facebook and Google derive their value from the activities and contributions of their users. They only provide the basic infrastructure for interaction. The value of crucial control points in extracting value was already demonstrated by Intel and Microsoft in the PC industry. It was a common observation that by controlling the chip architecture and the operating system software, respectively, this duo was earning monopoly profits, while the rest of the industry had to be content with commodity status. Today we are further along the way to what we might call the “Tom Sawyer economy.” In Mark Twain’s classic, Tom Sawyer, condemned by Aunt Polly to a day of hard labor, manages to inveigle neighborhood boys into paying him for the privilege of doing this work. While few businesses have attained this pinnacle of success, many have come close. For example, Angie’s List relies entirely on user-generated material, but charges for access to it. Among others in a similar situation is, of course, academic publishing.
Paul Krugman has started to focus on a related point, the massively increased role that rent extraction is playing in corporate profits, and its implications for the future of the economy:
A monopolist can, after all, be highly profitable yet see no good reason to expand its productive capacity. And Apple again provides a case in point: It is hugely profitable, yet it’s sitting on a giant pile of cash, which it evidently sees no need to reinvest in its business.
Or to put it differently, rising monopoly rents can and arguably have had the effect of simultaneously depressing both wages and the perceived return on investment.
You might suspect that this can’t be good for the broader economy, and you’d be right. If household income and hence household spending is held down because labor gets an ever-smaller share of national income, while corporations, despite soaring profits, have little incentive to invest, you have a recipe for persistently depressed demand.
He blogs about an initial model that shows:
What you have to imagine, then, is that some factor or combination of factors — a change in the intellectual property regime, the rise of too-big-to-fail financial institutions, a general shift toward winner-take-all markets in which network externalities give first movers a big advantage, etc. — ... [has raised] the profit share while actually reducing returns to both capital and labor.
The analogy with the "Big Deal" is obvious.

Aspects of this issue are popping up in the news as I write this. Bloomberg reports:
BC Partners Ltd., a London-based private-equity firm, agreed to buy German academic publisher Springer Science & Business Media GmbH for 3.3 billion euros ($4.4 billion) from EQT Partners AB and the Government of Singapore Investment Corp.
Springer Science, with more than 7,000 employees and sales of 981 million euros in 2012, publishes 2,000 magazines and 7,000 books every year on subjects including science and medicine.
In other words BC partners think Springer's publishing operation is worth 3.35 times revenue, and Springer thinks it is worth less. Why would BC think this? Apparently:
BC plans to focus the business on areas such as emerging markets and open-access publishing,
Springer has already been proliferating open-access journals, so that can't be a big part of the motivation. And emerging markets don't have huge library budgets to squeeze. My guess is that this is a typical private equity deal with the goal of cutting costs to make the numbers look better in the short term in order to seduce a greater fool to take the remains off their hands.

Taylor and Francis, a large for-profit publisher, recently published a white paper Facilitating access to free online resources: challenges and opportunities for the library community (PDF). Considering the source, librarians would be well-advised to take it with a pinch of salt. Among the take-away headlines T&F point to are:
  • 92% of librarians agree that free online resources are ‘very important’
  • Librarians feel they are well-placed to provide expertise in free content selection and discovery
  • 84% of respondents said that 10% or less of their time was currently devoted to indexing free online content
  • Key challenges for librarians relating to making free resources more discoverable within their institutions are: volume growth, unknown permanence, and difficulties relating to quality assessment
Among the conclusions I would point to are:
Librarians are seen by faculty as "purchasers of content" So their role relating to enhancing discoverability of free content and integrating it with paid-for content needs to be better promoted and developed; a particular challenge relating to this will be proving the return on investment of that effort.
Librarians have a critical role in helping their users spend less time searching and more time finding and reading content that they need for their research or studies. General search engines will be increasingly challenged to provide the level of filtering that will be required and librarians are well placed to develop methodologies and systems for the evaluation and presentation of a wider range of information resources tailored to the needs of a specific institution. However, support is required from a range of stakeholders in the scholarly information supply chain.
But, as Andrew points out, library users don't seem to agree with librarians and publishers about the inadequacy of general search tools:
the number of requests for reference assistance dropped by 66% in absolute terms, ... and thus by about 75% on a per-student basis.
Competing with Google to provide users with tools for finding information hasn't been very successful up to now; it isn't clear what has changed to make the prospects better. But Andrew shows that publishers have an interest in pushing librarians to focus on activities that the faculty don't value highly; that helps push library resources toward publishers.

1 comment:

  1. I don't know whether you're aware of the little book I did, The Big Deal and the Damage Done, looking at the effects of vastly increased serials spending on book (and other non-serial) and other spending in academic libraries on a sector-by-sector basis. It's a sad story.