books (Photo credit: brody4)

When I began to study the implications of patron-driven acquisitions (PDA) a couple years ago, I concluded that PDA services would put upward pressure on book prices as publishers sought to recover some of the sales and margin they would lose to the new demand-driven systems. While it’s still too early to determine longer-term trends in pricing, an outline of how pricing will evolve is forcing itself into view. But that view is not entirely clear, as the number of players in the supply chain and the availability of many titles from multiple sources make it very hard to determine what an individual book costs.

Libraries purchase books in a number of ways.  Some are purchased in what could be called the old-fashioned way — librarians reviewing individual offerings or responding to an inquiry by a faculty member; sometimes a librarian may simply happen upon a good review. In my experience, this is how most people, including career publishers, believe librarians work most or all of the time, but in fact most books are purchased through approval plans, where the library works with a vendor to determine the kinds of books that are appropriate for the institution. The vendor then sorts through its extensive inventory and presents only a subset of titles, which the library will either approve and purchase or decline. There are large administrative efficiencies in such plans (nothing is more expensive than title-by-title acquisitions), so libraries typically are rewarded for participation in approval plans with a small discount.

Books are also acquired in aggregations. These programs came into existence (inevitably, in my opinion) with the emergence of digital technology. The game here is to make books look more like the journals business, which is now mostly electronic. A digital aggregation of books can be sold like a digital aggregation of journals, aka a Big Deal. As I write this, several vendors are vying for the book-aggregation business, and some very large publishers (e.g., Springer and OUP) have aggregations of their own.  The benefits to libraries of such aggregations are clear — the efficiency of making a large purchase, a discount for buying so many books at one time, the affordances of digital content including full-text search, etc.  Publishers like aggregations because they make it possible to sell many books that might otherwise be passed over in approval plans. And that is why a discount is offered:  what is sacrificed in margin per title is (one hopes) more than offset by volume.

It may or may not be the case that aggregations will add to publishers’ sales to libraries. Short-term, they almost certainly will, but longer term there is the inherent problem of eroding course adoption sales. Let’s say you have a book on the backlist that reliably sells 500 copies a year for upper-level course adoptions. Now a digital copy of that book can be found in the library collection, where any number of students can check it out simultaneously.  How long before those 500 copies disappear into the aggregate library sale? I have seen figures that document this erosion in other segments of professional publishing, and if these figures were public, I doubt many publishers would participate in digital aggregations. Aggregations, in other words, are an administratively efficient way to sell weaker titles. That doesn’t mean that they don’t have value; the market is diverse.  It simply means that the strongest editorial programs will be the least likely to be included in such bundles.

With PDA (increasingly called DDA for demand-driven acquisitions) records for books that a library does not yet own are placed into the library’s catalogue. If a patron seeks to check out such a book, then an order is placed. For e-books the order is filled immediately; for print PDA programs, there is a delay until the book can be shipped ot the library. PDA allows libraries to hedge their bets on books that may be of interest (they got through the approval plan), but for which there may in fact be no express demand from faculty or students or that demand may take years to assert itself. Since the economic relationship between a library and a publisher is a zero-sum gain (when one wins, the other loses), the benefits of PDA to the library are likely to result in publishers fighting back, most likely in the form of higher prices.

So let’s make up some numbers to illustrate these different ways to acquire books. I emphasize that these are indeed made-up numbers and do not reflect actual pricing in the current or evolving marketplace.

Let’s suppose the list price of a book is $50. If a library is purchasing books one by one, this is what the library would expect to pay. While fewer and fewer books are purchased in this manner, the list price establishes a baseline against which all other prices are compared.

An approval plan is more efficient than purchasing books one at a time. Thus the approval plan vendor is likely to provide an incentive to the library for using such a plan. Let’s imagine then that the price of our book with a $50 list price is $45 under an approval plan.

The same title may be placed into an aggregation. So now we have 1,000 titles sold as a single package.  The average list price for individual titles in that package is $50, but libraries will not be willing to pay that amount without getting to cull certain titles. This results in the package being offered at a substantial discount; perhaps the entire package is made available for $30,000, which means that the imputed value of our hypothetical single title is now $30.

Now that title is placed into a PDA program. Publishers have multiple challenges here. Few publishers are large enough to support their own PDA programs, so decision-making is complicated by the need to work with vendors (who exact a slice). But the publishers will seek to get a higher price for a PDA title to offset the lost sales that are the basis of PDA (because the publishers’ lost sales are the libraries’ gain). So our book with a $50 list price may have a PDA price of $60.

It is not clear at this time whether publishers will be able to implement systems to enforce differential pricing for PDA and sales of the traditional (one book at a time) method, but they will try. If they are not successful, the higher price will be the only one.

So our $50 book can be purchased for $50, $45, $30, or $60. And we haven’t even gotten started yet.

There are several other variables that play a role here:

  • Format. It makes a big difference if a program is entirely digital or includes print editions.  Rationalizing pricing across formats is hard to do, and it becomes even harder when you may have the print edition available from one source, the e-book available from another.
  • Intermediaries. Most books are sold through middlemen, but there are exceptions, especially for ebooks.  The intermediaries impose their own constraints, which can add new variables to costs and pricing.
  • Unconventional sources. A company like Amazon may sell one book at a time (in print only, at least at this time), but may do so at a steep discount. Amazon may sell our $50 book for $40 and thus make the evaluation of all the other options that much harder.
  • Multiple sources.  What about the book that is available through three different aggregators and perhaps also available from Amazon and other online retailers and perhaps the publisher itself? Surely this will invite some form of discounting.
  • Short-term rentals.  An aspect of PDA is that it is not always about acquisition; it is often used for short-term rentals of a title. So to acquire our $50 book through a PDA program may cost $60, but to rent it for a day or two may cost only $10.
  • Usage rights.  Some titles may be for one user at a time, another may support multiple simultaneous users. Some titles may enable use in different ways (e.g., integration into a learning-management system), others may not. New use cases are invented daily. How to price these things?

The fact is that marketing books is now as complex as the marketing of any product, and it is much more complex than marketing journals. As an industry, book publishers have not caught up with this yet. Few publishers have the expertise to analyze all these variables.

Another way to look at this is to say that, broadly speaking, books and all other materials in the academic market really have two kinds of costs:  the cost for access to the content and the administrative cost of acquiring it.  What publishers, with the help of library vendors, have been able to do is to reduce the administrative cost to libraries and then turn around and increase the cost for content.  In effect, when libraries come up with more efficient workflows with the aid of their many suppliers, publishers turn around and raise the prices for the materials that are put through that workflow. This is how the Big Deal works:  a substantial collection from a single source, but at a whopping price. Publishers have effectively been hollowing out libraries, taking on more of the administrative tasks and leaving libraries with the hard-to-curate tasks that resist industrial engineering. Libraries get the overhead, publishers get the profit.

We should expect libraries to push back at this, and they are. The library-as-publisher is a popular meme for a good reason. Libraries will also seek ways to sort through the various and seemingly intentionally confusing offerings to get at just what, exactly, is the price of something. There is a real risk that publishers have pushed things too far. Librarians may decide to vote with their checkbooks and just say no.

Enhanced by Zemanta
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.


15 Thoughts on "What the Heck Does a Book Cost?"

“I have seen figures that document this erosion in other segments of professional publishing, and if these figures were public, I doubt many publishers would participate in digital aggregations.”

Not that I doubt your statement, but this is a pretty broad stroke statement which could be used to create FUD (Fear, Uncertainty and Doubt). I would like to see this sentence expanded to at least validate the sources and what frequency the numbers support the claim.

To the general theme of the article, there is an adage in sales: Where there is mystery; there is margin!

It is to the publishers advantage to keep it confusing and convoluted, whereas it is to the libraries advantage to understand the minutia. There is no question, the per unit cost will always be better with the bigger deal, what libraries need to figure out is how big a deal they can actually afford:

If you buy six dozen eggs at $0.03 per egg, but you can only ever eat a dozen, you may be justified in buying twelve $0.10 eggs…

The paperback challenge is huge. How many publishers can reliably predict which books will sell well for course adoption (unless the book was specifically written for that purpose to begin with)? I can cite any number of monographs, including revised dissertations, that I acquired for the two university presses I worked for that had startlingly large sales in paperback–in the tens of thousands–that no one predicted in advance. Now, with PDA, a publisher is obliged to make these guesses up front and decide whether to include titles in a PDA program. Those publishers who guess wrong risk losing a substantial share of their income stream. This is scarey stuff!

The challenge for libraries in becoming publishers is to carry the same level of prestige that university presses have acquired for purposes of tenure and promotion. Will they be able to take advantage of their parent university’s name in the same way that presses have? Time will tell . . . .

I always enjoy your thoughts on PDA. I think it is just too early to see any major impact on book prices brought about by PDA. Consider that most STM publishers book revenue is still coming from print. Spend a day out at the B&T warehouse in Sommerville NJ and you will see a very active print book shipping operation. In the STM world I believe that about 70% of the sales are still for print books. Most of the e-book sales to libraries comes from the traditional book suppliers or the new players now owned by ProQuest. I believe that most publishers may not even track the sales via PDA so it is not even on their radar. Here I do not include the University Presses as I don’t know how they are faring with e-books. Most STM publishers sell e-books through multiple channels but at this time the sales traffic is somewhere around 25% of their book sales.
Libraries are an important part of STM publishers monograph sales and publishers do monitor their print traffic and pay attention to their sales traffic through the various channels. I am not sure how much deeper the monitoring has developed. Is PDA sales even a sales category with most STM publishers?

What I do know is that the monograph sales traffic to an individual library has dropped where the library has implemented a PDA program. The group feeling the most impact is the book suppliers who are seeing reduced sales. The book dealers should be the first group to modify their pricing for services. In the past a library spent $100,000 on print books and now that amount has often dropped to $30,000.
While it is hard to generalize, most STM publishers make nearly all of their revenue on the sales of site licenses. Books are a very distant second and far less profitable category.

Two thoughts.

First, is there really a problem here? Packets of cornflakes cost different amount in different retailers. Commodities such as flight tickets and hotel rooms, sold within a yield management programme, vary even more. When you write “So our $50 book can be purchased for $50, $45, $30, or $60”, you seem in effect to be reifying the $50 price — as if that’s somehow the true price and the others are distortions of it. Why not just accept that different circumstances produce different price points?

Secondly, I think there’s more than one publisher perspective on PDA. My ininitial response when I learnt of the idea was that it would help me because it would reward quality – the more recommended, cited, etc. a book becomese, the more that triggers purchases. The publisher it would disadvantage would be the large-but-not-very-good publisher, resting on its laurels relying on scale to produce sales through approval plans.

I also thought that PDA would reward good backlist publishing and so advantage publishers (such as ourselves) who commission books designed to have long-term appeal.

So my initial response was positive – and it remains so. If you publish good quality backlist sellers, the problem of recovering sales and margins doesn’t arise.

So, how do you decide what books to put in a PDA program, with regard to the effects on paperback sales?

I’m somewhat reluctant to engage in the paper/print debate, only because – as I dare say you know as well as i do, Sandy – once it gets started, it tends to elicit the same-old views over and over. I don’t have an original viewpoint on that. My own angle is that,from our own purely commercial view, (a) e-books don’t drive out p-books and (b) I don’t care if that do. Not least because we price the e-book higher than the paperback.

I beg to differ. As Joe says, when a library buys an e-book, it is available to the entire population of the campus, and it would make no sense for a teacher to ask students to buy a paperback of a book that is available for free to students through the library. Only if your sale were to impose a restriction of one user at a time would you avoid the consequence of wiping out your paperback sale to that market. This is a huge dilemma for anyone who sells e-books through PDA or through aggregations like Project Muse.

Thanks, Sandy.You’ll note from my language above that I wasn’t advancing a principle for the industry as a whole, just reporting our own strategy.

Perhaps there’s a British/American distinction in viewpoint: UK universities will only rarely use a monograph as a course text, whereas American institutions are more likely to.

Also, we find that many university libraries like to buy BOTH the e-book and the p-book. There are many reasons for this. To some extent, the e-book still acts as a discovery tool for the p-book.

In addition, some suppliers use DRM to limit/control the usage (I was reluctant to add that point, though, because that’s another (very boring!) can of worms).

The fundamental point is, I think the best way to publish monographs is to do so with a business model that works on a couple of hundred sales per title. If university presses are struggling to make their own models work, why not just shift models?

A further option, already adopted by some commercial presses, is to establish a cloud library. A university press with a backlist and forward programme of any size would be in a good position to do so, though of course that would require them to be (a) tech-savvy and (b) ready to innovate – not qualities I think of characterising the strengths of that sector!

It’s not the loss of a single print copy that is the problem. It’s the loss of multiple copies of paperbacks used in courses. paperback sales for American university presses generally account for some 40% of overall revenue, and except for a few titles (like regional books) that have trade sales, the vast majority of these sales are for course adoption. Any significant diminution of that market through PDA will put a press in financial jeopardy. Monographs in the U.S. do sometimes sell in very large numbers. Revised dissertations like Susan Okin’s WOMEN IN WESTERN POLITICAL THOUGHT and Peter Evans’s DEPENDENT DEVELOPMENT, both of which i acquired at Princeton U.P., each sold well over 30,000 copies. If those books had been published today and sold as e-books to libraries as part of a JSTOR package or individually with no restriction on multiple usage, the loss of income would be very substantial indeed.

This seems to me to boil down to a choice between two options: persisting with yesteryear’s model and fretting about how the world no longer accommodates it; or devising a business model for the world of today. If market change truly puts presses “in jeopardy”, isn’t that just another way of saying they’re too lethargic to change their business models? Frankly, I can’t see why anyone in 2013 would want to spend their time commissioning revised dissertations. Why would anyone — editor, publisher, librarian, retailer, reader – get out of bed for that?

Anthony, the reality of the system is that junior professors must try to get their dissertations published in revised form in order to earn tenure and advance in their careers. Most top universities require two published books now for tenure (in the humanities), and unless a junior faculty member uses the dissertation as the basis for one of them, it is a huge challenge to meet that goal in the six years they have to do it. This has nothing to do with publishers commissioning revised dissertations; it is simply the way the system works. So, university presses changing their business models is not a solution since these presses are mandated to serve the present system. If they do not, their parent universities will see no reason to continue supporting them, and indeed several presses have been closed in recent years. That said, the examples I gave of two revised dissertations obviously show that some of them can be highly successful, both commercially and critically (these two books are considered pioneering in their respective fields).

It is important, surely, to think logically about all this. And that requires that we don’t elide two distinct problems. First, we were discussing the economic issue of how publishers should respond to changes in the market, especially concerning PDA and price. Now we have slipped to discussing how dissertations can be published. There’s no reason to assume that both problems can be solved, still less that a single solution can be found for both. They are distinct.

The root of the latter problem – the need for authors to get dissertations published in order to obtain tenure – has nothing to do with Joseph’s post. It has to do with universities’ stipulations regarding tenure. Those stipulations now look rather dated and, I suggest, plain silly, the solution lies in universities hauling themselves into the 21st century and doing a little thinking. Most doctoral theses do not make good books and they can in any case now be disseminated via repositories and online databases.

I should note here a problem with your argument, which is that you use the present tense ‘can’ about books published a third of a century ago. A similar,and similarly weak, argument could be made in favour of, say, sustaining the typewriter industry. That was then, this is now.

A further point: as I said above, it’s possible to operate profitable publishing models based on a couple of hundred sales of monographs. But really the monographs concerned need to be more than warmed up theses. And it may be that, because of lethargy and inefficiency, some university presses will fail to implement such models. But I see no reason to agonise over the fact that companies who persist in producing stuff for which there is insufficient demand will fail – a point I have argued in more detail elsewhere (

In short, university presses face a choice: either publish strong content (including monographs) using an efficient model suited to today’s market; or fail. Where’s the problem in that?

What kind of model do you think is appropriate for the times, Anthony? Presses like the one I headed at Penn State have been experimenting for nearly a decade now with alternative models (see Penn state’s “open access” monograph series in romance studies), but I don’t think anyone would yet say that we have found one that really works and is sustainable long-term (except, perhaps, one based on endowment, which is what the new press at Amherst will be using). You say that it’s possible to have a model that works with only a couple of hundred sales. Well, yes, that’s the model that European academic presses have been using for many decades now. But the question to ask about that model is how well it really serves scholarship–not very well, in my view, because it limits access to only those who can visit the few large university libraries that buy the couple of hundred copies. The prices are so high that individuals can’t afford them and they can’t be assigned for courses. That’s why the possible PDA threat to the paperback market is so alarming. It will help with the problem of access, but it could end a very important revenue stream that has allowed American presses to publish hardbacks at prices much lower than what European presses charge. This is where the argument connects up with what Joe is saying here. You eliminate the paperback revenue, and you’ll need to charge more for the hardback (or e-book).

Comments are closed.