Bite Sized!
Image by thehoneybunny via Flickr

Over the past couple of years I’ve been writing and presenting pretty relentlessly on the topic of patron-driven acquisition (PDA) in research libraries, arguing that in a predominantly online information environment it no longer fundamentally makes sense for most academic libraries to build large permanent collections based on librarians’ speculations about patrons’ future needs. For one thing, our speculations are very often wrong; for another, the combination of shrinking budgets and relentlessly increasing prices (especially for scholarly journals) makes traditional collection practices decreasingly feasible from a purely economic standpoint — we just can’t afford to keep building collections the way we have in the past.

PDA is built on a deceptively simple premise: in a largely digital information environment, it’s increasingly possible to let library users find and identify desired documents prior to the library’s purchase of them, and for the library to pay only for what its patrons find and actually use. When a patron’s use of an ebook or journal article passes a certain agreed-upon threshold (a certain number of ebook pages read, for example, or the download of a complete article) the library is charged, the document acquired, and the patron never knows that the document was not part of the “collection” to begin with.  Such an arrangement has the potential to be enormously liberating for library users, and to solve one of the library’s longstanding and fundamental problems: the fact that traditional “just-in-case” collections give patrons access to only a tiny (and inconsistently relevant) sliver of the population of documents that are actually available for use.

But like all acquisition and access models, PDA is imperfect, its manifestations are numerous and to some degree chaotic, and its availability raises lots and lots of questions, many of which I find myself trying to answer during the Q&A segments at the ends of my presentations. Several questions arise repeatedly, which suggests to me that there’s broad interest in answers to them. Here are some of those questions, with my attempts at responses.

Q: Isn’t all collection development patron-driven? After all, librarians have always striven to understand patrons’ needs and take them into account when selecting materials for the collection.

A:  In academic libraries, we’ve gotten very good at understanding our patrons’ needs in the aggregate: we know the curriculum and we know our faculties and their areas of research interest, and that knowledge has always guided our collecting strategies.  This means, for example, that it’s possible for me to know that my faculty needs good books on high-energy physics. The problem is that it’s not possible to buy “books on high-energy physics.” It’s only possible to buy specific books on high-energy physics, which necessarily entails not buying other books on high-energy physics, and my ability to predict which exact books on high-energy physics my patrons will need and use is very limited. Remember that the purpose of the collection isn’t to be a great collection; it’s to connect patrons with exactly what they need (see further discussion of the collection’s purpose below).

Q: Is it really our job to just “give the people what they want”?

A:  Where “the people” are the students and faculty members we’re being paid to serve, and where “what they want” are resources that support their scholarly work, then I think the short answer to this question is “Yes, that’s exactly our job.” The obvious corollary to this question is “But what if all they want is comic books and genre fiction?” The answer to that is: while PDA can offer a technically unlimited array of options to our patrons, fiscal reality makes some limitations inevitable. We may have to filter the options we show them; depending on our institutional mission, we might exclude romance fiction or travel guides from the range of options, for example (while leaving the door open to special requests from researchers working in popular culture or related fields). As long as resources are limited, no access model will be perfect.

Q: How will PDA help me save money?

A:  It probably won’t. And to my mind, that’s fine. I have no expectation that PDA will allow me to spend less money on books and articles; I only expect it to help ensure that all the money I spend will go to materials that my patrons actually need. Don’t get me wrong; I love to save money. But that’s not why I do PDA.

Q: How can you say that any purchase is a “waste of money”? You yourself said that we never know perfectly what our patrons are going to need in the future.

A: That’s true, but since it’s true of absolutely any document—there is no document about which you can’t say “someday someone might need access to this”—it doesn’t help me make good use of limited resources. Unless I have an unlimited budget, I have to choose between multiple potentially-useful documents, and that means I have to try to pick the ones that are most likely to be used. And my patrons can do that much more effectively than I can.

Q: Won’t PDA hurt publishers?

A: Yes, and that’s unfortunate. I mean that sincerely. But we have a problem, and the problem is that the current publishing marketplace evolved in an environment in which library customers had no choice but to buy lots of books and articles that they didn’t need, because that was the only way to guarantee access to books and articles that they did need. That’s not a criticism of publishers; it’s a criticism of the print-based information environment. PDA has the potential to make the information marketplace much more efficient and rational, which means that publishers will go out of business to the degree that they relied (however unintentionally) on marketplace inefficiencies to keep them in business. I wish that weren’t the case, but it is.

Q: PDA always seems to be about books. Doesn’t the same principle apply to journals?

A: Absolutely. The journal subscription is a fundamentally irrational way to buy access to articles; it’s a way of buying lots of content you don’t need in order to ensure access to some content that you do need (while excluding other big batches of content, which may also contain articles you need). PDA principles apply to journal content at the article level: ideally, libraries should expose huge and comprehensive “collections” of unacquired journal articles to their patrons, and buy only those that their patrons actually download. A certain number of paid downloads should result in permanent sitewide access (on the assumption that multiple uses of a single article demonstrates broad need on that campus). The details of such an arrangement would have to be a matter of negotiation with the publisher; the most important detail, obviously, will be price per article.

Q: How do you control your spending in a PDA environment?

A: This is a huge issue, and it’s one that is being handled in a variety of ways from both sides of the supply-demand divide. One possibility is to use a “risk pool” approach as a throttle mechanism: if the rate of purchase is outstripping budget, then the number of books offered for PDA is reduced until demand slows. On the supply side, publishers and aggregators generally offer management tools that make it relatively easy to track usage activity in real time, and to tighten or relax access as needed. In the article environment, it might make sense simply to ration the number of permitted downloads.  We should expect that publishers and aggregators will compete with each other, in part, on the basis of their ability to help us manage spending, so we should probably not expect a one-size-fits-all solution to emerge.

Q: My library subscribes to several comprehensive journal packages from major science publishers. The pricing for these packages is advantageous enough that buying the articles one-by-one would result in a much higher expenditure than simply sticking with the Big Deal does. Surely a PDA approach doesn’t make sense in my case.

A: It’s true that a Big Deal will often result in very low per-download costs. But that doesn’t mean that it makes sense to buy a huge batch of articles, only some of which will get used; it means only that the publisher has set pricing in such a way as to discourage per-article purchasing and encourage bulk purchasing. Such arrangements may be attractive in the short run, but they are objectively unsustainable in the long run. The Big Deal was not (pace Derk Haank) a solution to the serials pricing crisis — it was only a way of kicking the can of unsustainable pricing down the road for a few years.

Q: How can you build a great collection if you only buy what today’s patrons want? Won’t the result be a rather haphazard, possibly even incoherent pile of content rather than a well-rounded, intelligently shaped research collection?

A:  This goes to a fundamental philosophical issue on which I and many of my colleagues disagree. I believe it’s essential to the survival of academic libraries that we learn to see the collection as a means, not an end. By this reasoning, an arguably fine (i.e. high-quality, well-organized, and reasonably comprehensive) collection that fails to meet the actual real-life needs of the scholarly population it is supposed to serve is not a “good” collection in any meaningful sense. In fact, I’m not sure we can safely assume that building collections is still the right way to go about meeting patrons’ information needs at all. While it probably makes sense for a few richly-funded research libraries to build and maintain huge and comprehensive collections (I call these the Monument to Western Civilization collections), for the vast majority of academic libraries that kind of collecting has always been impossible and is now increasingly irrational.

Enhanced by Zemanta
Rick Anderson

Rick Anderson

Rick Anderson is Associate Dean for Collections and Scholarly Communication in the J. Willard Marriott Library at the University of Utah. He speaks and writes regularly on issues related to libraries, scholarly communication, and higher education, and has served as president of NASIG and of the Society for Scholarly Publishing.

View All Posts by Rick Anderson


55 Thoughts on "What Patron-Driven Acquisition (PDA) Does and Doesn't Mean: An FAQ"

My impression was that most large and small journal publishers allow individuals to scan unsubscribed e-journals current and archival “Table of Contents” and allow purchase of single articles based on the abstract or even (in the case of Springer) the entire first page. Doesn’t this amount to “Patron Drive Acquisition” already?

Yes: technically speaking, the model does indeed exist already for articles. What doesn’t yet exist is a sustainable pricing regime. At the moment, libraries have to choose between buying individual articles at extortionate prices and buying batches of articles at somewhat lower but still unsustainable prices. Neither of these will work in the long run; apart from the basic irrationality of the batch (i.e. subscription) model, the numbers just don’t work.

If one takes as a baseline assumption that we’ll have relatively flat or insufficiently expanding library budgets, then the faculty will not be sufficiently supported in purchasing individual articles, nor on-going electronic subscriptions, for their research needs. If libraries don’t or won’t or can’t expand their journal or individual article budgets, but international research output keeps rising, what do you predict as probably patterns for access to research over the next 2 or 3 years? Let us assume the best research is still in the best traditional (and non-OA) journals, for argument’s sake.


What you’re describing is a pricing problem. We can’t afford to buy all the articles that our patrons need, so we have to ration access; the higher the prices, the more strictly we have to ration. Of course, we’ve always rationed access — it’s just that traditionally we’ve done it at the journal title level, which has meant buying lots of articles our researchers didn’t need along with the ones they did, while also failing to buy lots of articles they did need.

I have a hard time predicting future patterns of access, because I don’t know how publishers are going to react to the wave of cancellations that has already begun and will only increase over the coming few years. Faculty reactions will have an impact too, and their impact will be especially interesting because they act on both the supply and demand sides. This coming year, for example, my library will have to cancel roughly $300,000 worth of journal and database subscriptions, and I anticipate strong faculty response. But we have little choice; the book budget has been eviscerated and really can’t be cut any further. Will faculty get angry at us for cutting, or at publishers for raising prices at unsustainable rates? And what will the faculty do with their anger? We’re on the threshold of very interesting times, I think. I wish I could predict the upshot better.

I take it then that article PDA requires some sort of deal with each publisher, making their articles affordable. Are the publishers doing this? Does PDA for articles even exist? It sounds very elaborate.

PDA for articles does exist (libraries usually call it “Document Delivery”), and the elaborateness of the models varies from publisher to publisher. The biggest barrier to systematic uptake of such arrangements is retail price: most science publishers charge $40 or so for a copy of an individual article, and obviously that’s not a sustainable price for a library serving tens of thousands of students and a couple of thousand faculty. And that’s part of the point: publishers would prefer that libraries pay up front at the beginning of the year for access to a large package of articles (i.e. a journal subscription) rather than buy only those articles that are actually needed by patrons. It’s a completely understandable preference from the publisher’s point of view, but, for the reasons outlined above, the journal subscription is a terrible way to give students and researchers to the articles they need.

I would say that $40 articles does not constitute article PDA, so while the technology may exist the functioning library system itself does not.

David, I’m just curious: why is price the criterion that determines whether or not you consider an acquisition model to be patron-driven?

  • Rick Anderson
  • May 31, 2011, 10:54 PM

Rick, I don’t understand your question so we must be talking past each other. My basic point is that article PDA can’t work at $40/article.

  • David Wojick
  • Jun 1, 2011, 7:40 AM

PDA for articles also exists in a new program that the CCC has launched called “Get It Now,” which is being used successfully now in the University of California system and provides much faster delivery times at lower cost than traditional ILL.

> Rick, I don’t understand your question
> so we must be talking past each other.
> My basic point is that article PDA can’t
> work at $40/article.

Your comment was that “$40 articles does not constitute article PDA,” so it sounded to me like you were saying that the patron-driven acquisition of expensive articles (as with traditional document delivery) isn’t really patron-driven. But it sounds like you’re really just saying that such pricing is unsustainable in a PDA environment, so we’re actually in agreement. Carry on!

something like the table of content (TOC, SDI or current content model) isnt it? we did that when back in the 80s and 90s becoz we want our user to select articles from many journals that the library subscribes…

What is the use of the “Big Deals”?

80/20? Most journals in these packages are not used or used once or twice.

Isn’t this publisher driven acquisition?


The Big Deal is the exact opposite of PDA: it involves the just-in-case acquisition of an enormous package of articles on the assumption (usually correct) that enough of them will be used to make it a cost-effective purchase. The problem is that even though the per-article cost often ends up being quite low, the Big Deal is not a sustainable model; eventually it will crowd out all other purchases, and you’ll be left with lots of articles you don’t need and no further ability to buy articles from other publishers that you do need.

While I think there are some merits to PDA, I don’t think that it is any more “sustainable” than other models. Indeed, I think your use of “sustainable” is a red-herring.

A large move to PDA will induce publishers to limit sharing of books and journal articles, requiring the library to purchase that book or article over and over and over again. The net result will be an overall reduction in the diversity of access.

As long as the corpus of literature is growing beyond what institutions are willing to support their libraries in terms of their budget, any solution will be “unsustainable.” It may be better to call PDA more efficient, less labor intensive, or more sensitive to patron’s wishes, just not more “sustainable.”

Hi, Phil —

Look again — nowhere did I characterize PDA as a sustainable model in itself. That’s because sustainability, as you point out in your comment, is not a function of model, but of pricing. Irrational models can be priced low enough to make them sustainable, but that doesn’t mean that they make sense at a fundamental level.

Where you suggest that I characterize PDA as “more efficient, less labor intensive, or more sensitive to patrons’ wishes,” I would prefer simply to say that it’s “more rational.” In a perfectly PDA-driven environment, every dollar spent would buy something a patron actually needs. To my mind, that makes the model rational (even if it isn’t, for example, necessarily less labor-intensive). Now to make that _rational_ model _sustainable_, we have to figure out the right pricing.

You suggest that a PDA environment will induce publishers to find ways to limit sharing, requiring libraries to purchase multiple copies. That’s a pricing issue, and a point for negotiation. An important part of negotiating any PDA arrangement is coming to an agreement on how many purchases of a single document (book or article) will result in the permanent acquisition of that document (meaning simply that additional uses will not generate additional charges). In this as in all other aspects of library acquisitions, sustainability comes down to pricing.

I don’t think you can simply separate pricing models and pricing. Nor can you apply the terms “rational” or “irrational” to a pricing model. Rational models have to do with internal consistency. Thus, there is nothing less “rational” about a bundle deal than a per-article payment model. Indeed, a libertarian argument that we should shut libraries and redistribute its collection funds to faculty and students to purchase their own materials online is also a rational model. In the end, your decision-making is ultimately based on rationing a limited amount of resources in your community. How you do it is not a mater of what is “rational” or “irrational” but your values and priorities.

My argument is based on the premise that it is less rational for a library to expend scarce resources in buying a document that no patron wants, and that it is more rational to expend those resources in buying a document that at least one patron does want. If you disagree with that premise, then we’re doomed to talk past each other on this issue.

(If all you disagree with is my colloquial use of the term “rational,” then feel free to substitute “sensible” or “sane” or “logical” or whatever term you prefer. My point is that some ways of allocating of scarce resources make more sense than others, given the goal of meeting patrons’ information needs.)

I take your point, but language here is important. If you are arguing that there are some better ways to allocate resources than your language needs to be pragmatic in nature and also needs to take context into consideration. This means you need to avoid using terms like “rational” and “sustainable” when describing PDA.

  • Phil Davis
  • May 31, 2011, 4:20 PM

I take your point, but language here is important. If you are arguing that there are some better ways to allocate resources than your language needs to be pragmatic in nature and also needs to take context into consideration. This means you need to avoid using terms like “rational” and “sustainable” when describing PDA.

I agree that language is important, and I disagree that my use of either “rational” or “unsustainable” in this context lacks either pragmatism of consideration of context. If you don’t like the word “rational” as applied to the sensible allocation of scarce resources, then fine — although I think “rational” is perfectly accurate, there are other words that work equally well for my purposes.

But I have to push back on “unsustainable,” especially since you’re insisting on the importance of both pragmatism and context. Here’s the context: an environment in which materials budgets are generally either flat or shrinking, and in which the average annual price increase for a scientific journals is around roughly 9%. The Big Deal carves out a large chunk of the library’s materials budget and fences it off, so that the package’s useless content must continue to be bought (at an annually increasing price) from year to year. This arrangement may be pragmatic in the short run (because along with the useless content, one also gets access to useful content at an advantageous price) but it is objectively unsustainable in the long run, for the simple reason that it cannot be sustained. First, the Big Deal forces the library to stop buying access to other useful content; eventually, the Big Deal becomes unaffordable itself, unless the library’s annual budget increases at least match the annual price increase of the Big Deal. The word “unsustainable” in this context applies in about as purely pragmatic a sense as one could imagine.

Interesting article with a very narrow focus on PDA information distribution. The plaque remaining in our informational arteries is printing, storage and dissemination controlled by publishers who decide what is relevant to publish based on PDA informed orders from libraries. That concept is now in free fall and will come to an end very soon. Quality control and refinement of information will take a big hit but those who provide product refinement services will be individuals setting up their own categorical portals and compensation models. The overlooked aspect of any article on PDA is global informational access that is razing the language barriers to provide broader subject insight. Discussions of PDA are tangential to the main topic of information dissemination through the constraining prism of libraries as viable institutions.

Hi, Ron —

I confess that I couldn’t completely follow your comment. It seems like you’re saying I should have written about something other than PDA, which you seem to feel is both too narrow and also tangential to “the main topic,” but I’m not entirely clear what you mean by “information dissemination through the constraining prism of libraries as viable institutions.”

Switching from subscriptions to PDA-driven per-article purchasing could curtail what gets published. Some editorial boards already use citation data to determine what content to publish as a way of driving up their Impact Factors. Research that is of interest to a smaller number of people could find itself homeless if article sales do not justify publication. Some of these articles turn out to be important years after publication. A PDA model could prevent them from ever being published.

That’s true. But let’s consider what we have now: a system that encourages and facilitates the publication (at great expense) of articles many of which are purchased (at further great expense) and then never used by the people for whom they were purchased. The current model is massively inefficient, but there are many who benefit directly from those inefficiencies, whether that is their intent or not. If we move to a system that is more rational (by which I mean “involves the throwing of less money down the toilet”), those who benefited from the old system will certainly lose some of those benefits. My belief is that the overall benefit to researchers will be worth the cost — and I say that as an author as well as a librarian.

In 2004, Ebook Library introduced their PDA model, called Non-Linear Lending (TM). My job at that time was to persuade librarians to try it. The results of those early pilots were eye-opening. I remember, in particular, the head of a large community college library calling me to exclaim that one of the first-used titles was from an STM publisher in physics. Not something on his short list for purchasing, but real data about student interest — and an opportunity for publishers to have their titles discovered. (in a PDA world, discovery continues to be a critical driver of sales).

Alix, you’ve hit on a very important point about PDA: it’s not just about increasing the efficiency of acquisition and avoiding spending money on the wrong things. It also has the potential to get libraries back into the discovery game. The library catalog does a very poor job of answering the question “Is there such a thing as a book on [X]?”, because the catalog interrogates such a limited population of documents. But if the library is brokering a PDA experience that allows users to search many millions of documents rather than only those that are held in the library’s collection, the library suddenly becomes a player again in terms of discovery. That’s one of the more exciting things about PDA, I think.

Rick –

To extend the concept a bit, here’s my take on patron-informed/driven acquisition in e-resources, from the current issue of Library Journal: I’d love to hear your thoughts on it as/if time allows.

Peter McC

Great piece, Peter — and I like the model you propose for access to a reference database. Where the service being provided is access to smallish bits of information that will mainly be looked at rather than larger chunks of information (like articles) copies of which are more likely to be taken away than read at a sitting, or even larger ones (like ebooks) that are likely to be interacted with in a more extensive way online, then I think it makes good sense to impose microcharges by the use. As you point out, the devil is in the details — but as a general model, I think what you propose is eminently sensible.

Are you guys considering an approach like this for

I see this as something that Summon, EDS, Primo Central, or others would implement. So I can’t make it happen unless they do, but I think it’s an excellent next step for any or all of them, for lots of reasons.

I’d love to apply it to — I believe that while many librarians think their patrons wouldn’t use the database, a pay-per-use model would let them test it with little risk. And, I think, most institutions would discover that the database *would* get used. Maybe not always enough to justify a direct subscription, but certainly enough to have it available on a pay-per-use model.

PDA sounds like an opportunity for a new kind of aggregator. One who can cut deals with various major publishers for low priced articles on demand,then shop that deal to libraries, then process the traffic. Is there an entrepreneur in the house?

It’s called the “Get It Now” program, and the entrepreneur offering it is the CCC.


In a sense, some aggregators already exist. EBSCO aggregates some major journal publishers, but with an embargo period; the service “DeepDvye” also aggregates some journal publishers with an “iTunes” type rental service for journal articles, ranging from 99 cents to $2.99 per article. If you haven’t seen that service, go to .

Best wishes,


Rental is an interesting variation on PDA. With purchase I can see a buildup of not-needed articles left over from old projects.

Rick and I have engaged in a dialogue over PDA that appears in the June issue of Against the Grain. Here I’d just like to observe that one can push PDA a step further, as Eric Hellman has suggested, and create acquisitions consortia whose purchasing power would be such as to put them in a position not just to buy what publishers already offer but to serve as “subscribers” in the 18th-century sense whereby a prospectus for a book was floated and the work only published if sufficient demand for it existed. This could be a further rationalization of the model Rick urges, sparing publishers the expense of speculating on producing books no one wants. This is one possible way in which university presses might be able to survive and husband their resources. If no one wants a book, the question should be not just why buy it but also why publish it in the first place?

This continuing discussion of Patron-Driven Acquisition is welcome and instructive. Rick’s FAQ offers his often compelling responses to a number of issues. Other matters may warrant attention as well, among them questions of coverage, memory, and community.

Coverage. PDA builds from a pre-existing base file of bibliographic records from which users make their choices. It reflects a universe of materials that has been defined beforehand—a universe today pretty much limited to the English-language scholarly and trade publications that a particular aggregator can provide. Materials not meeting these criteria will not be represented in the files that users see: for practical purposes, they simply don’t exist. While the ideal model for PDA suggests unlimited user-driven access to everything of potential interest, the reality falls significantly short. Foreign-language materials, items from regions with weak book trades and information marketplaces, primary sources, and analog materials are all at risk. Incomplete coverage has always compromised expansive collections aspirations. PDA would reframe this inevitable limitation as an inherent virtue. The argument would be stronger if users were assured of a more complete beginning universe from which to choose.

Memory. PDA speaks to the satisfaction of users’ immediate needs. Research libraries have historically also perceived preservation as a core function. Together, research libraries have accepted an implicit responsibility to keep the records of scholarship and of human expression alive for future generations. Broad questions of heritage and memory are very much in play. PDA, as currently framed, elides preservation as a concern even for the materials that patrons select. Can preservation and memory reasonably be relegated to the marginalizing rubric of “Monument[s] to Western Civilization”?

Community. PDA addresses current users’ immediate needs, at this moment within the universe of English-language scholarly and trade publications that are available in digital format. Few colleges and universities can expect to meet all user needs from so narrow a base: interlibrary loan and document delivery services are unmentioned but essential service corollaries. But reliance on ILL presumes (1) that someone, somewhere has acquired the low-use materials suddenly in demand, and (2) that the “someone” in question is willing to make these low-use materials available. Our system of libraries has heretofore assumed interdependence, as even the largest libraries have needed to borrow some materials. Broad-gauged reciprocity has then supported an ethos of sharing at the lowest possible cost. The PDA model, which is silent with regard to libraries’ interdependencies, raises the prospect of massive reliance on ILL to close local collection gaps. If a few collection-intensive libraries become de facto back-ups for a great many others, reciprocity will go by the boards. One possible result could be ILL fees that fully reflect collecting libraries’ full costs, which might run into the hundreds of dollars per transaction. The nature of our research library community would change accordingly—perhaps through more effective shared initiatives, perhaps through version to a bibliographic state of nature.

PDA carries many compelling benefits for users. It may also allow individual libraries to operate more efficiently. The consequences for broader communities of libraries and scholars seem far less certain.


Thanks very much for your (as always) thoughtful comments on a complex issue. I suspect that it will generate many well-deserved “likes,” and my response will probably generate the opposite. But I’m going to soldier on anyway. Let me respond to some of your points individually:

1. Coverage — You’re right, of course, to point out that a PDA model will never draw on the entire universe of existing documents. Even if it could do so technically, no library is funded sufficiently to offer its patrons everything they might possibly want. But let’s look at what the alternative model has been up until now: access to a radically more-limited universe of options (i.e., the traditional library collection). As I said in my posting, PDA does not solve the access problem completely — and it most certainly does not “frame… [that] limitation as an inherent virtue”; at best, I believe PDA can move us in what seems to me the right direction: away from an artificially tiny array of documents (as selected by librarians based on speculation) and towards a radically expanded universe of documents, including the kinds of long-tail documents you mention, virtually none of which are available to the vast majority of library users served by traditional library collections. I think about it this way: my goal is comprehensive access for all of my patrons to everything they need, in the moment that they need it, in the format that works best for them. Such a goal is unachievable, of course, at least for now. However, to stay pointed at that goal is, I think, to ensure that my library is moving in the right direction. I think that PDA, however imperfect and embryonic it is at this point, is an important step in the right direction when it comes to access.

2. Memory — While I agree in general terms with your points here, I would qualify them. Where you say “Research libraries have historically… perceived preservation as a core function,” I think I would say “Research libraries have, in varying degrees, devoted resources to the preservation of certain specific categories of material.” You are right that PDA is about meeting immediate needs, and that it does nothing to meet the goals of preserving heritage and memory; another way of putting it is that PDA is about doing what general collections do, only better, and not about doing what special collections do. The general, circulating collections of my library (and, I think it’s safe to say, of most other research libraries) do not serve a significant preservation function. Materials are added to and withdrawn from the general collections on a more-or-less regular basis; damaged copies are discarded or replaced, sometimes with different editions and printings; books are lost and stolen and only replaced if prior usage justifies it. This part of the library is designed to meet immediate needs, not to act as a memory bed for our culture, and it is this part of the library that I believe cries out for a more rational acquisition model. PDA is about access, not preservation.

3. Community — Here I want to make two points. The first is a quote from Cliff Lynch: “The purpose of collaboration is not collaboration.” I would tweak that line for our purposes by saying that “the purpose of a library community is not to be a community” or, in other words, sharing between libraries is a means, not an end. The purpose of the library community is to serve library users. This brings me to my second point: in your comments you conflate ILL and DocDel in a way that I think elides (I too love that word) a vital distinction. DocDel is a rudimentary and imperfect form of patron-driven acquisition — a way of giving the patron exactly what she wants when she tells us she wants it. ILL, however, is completely different; it’s the library’s attempt to make up for its failure to speculate accurately about a future patron need. With ILL we say to patrons “Sorry, we failed to guess that you would want that document. Sit tight and we’ll see if we can borrow a copy from a different library. Hopefully it’s not a rare book or a very new publication or a reference work.” To say that PDA would undermine our ILL regime is, I think, much like saying that increasing our production of renewable energy would undermine our oil-pumping regime — true enough, but isn’t that the whole point? To my mind, ILL is not a core function of libraries and a salutary example of community sharing; it’s a necessary evil that we should not put up with any longer than we have to. I think we should be working towards an information environment in which ILL is unnecessary. Again, we’ll likely never reach that perfect environment, but in the meantime let’s not make the mistake of confusing onerous necessity with virtue.

However, you point out a very real problem with the idea of asking a few deeply-funded libraries act as a culture’s documentary memory and letting less-wealthy libraries depend on them for access to low-use documents. In the past, access to low-demand documents has been essentially crowdsourced — the University of Utah may not have access to Document X, and even Harvard or the Library of Congress might not have it, but thanks to our traditional collection-development practices the chances are good that some library, somewhere, does own it and will be willing to lend it. That’s a real benefit of the old system. But as with all propositions, a rational assessment requires us to look at both the benefits _and_ the costs. It seems clear to me (though I realize there are smart people who disagree) that the costs of providing access in this way greatly outweigh the benefits.

Let me propose a scenario. Let’s suppose that Document X exists in only two copies, and that it’s a document that constitutes an important and preservation-worthy piece of our intellectual heritage. This implies two imperatives: an access imperative and a preservation imperative. Traditionally, we would meet the access imperative by shuttling the book around between libraries upon request — a practice that not only provides lousy access, but also threatens the document’s physical integrity, thus failing at least partially on both counts. Now suppose instead that we scan the document and make it available for immediate digital delivery (with payments going to rightsholders, as needed) on a PDA basis through libraries. Access is now extremely broad (though technically not universal) and potentially instantaneous, and the document’s preservation is no longer threatened by the distribution method. If the requesting library has print-on-demand capabilities, it can even provide the patron with a physical copy of the digitized text. “But wait,” I hear you cry. “What if the patron genuinely needs access to the original physical document itself, not just a digital image or printout of it?” In that case, we’re no worse off than we were to begin with; traditional ILL remains an option where needed, but hopefully the need for it will be greatly lessened. And a PDA system, while not solving either the access problem or the preservation problem completely, has moved us much further in the direction towards a solution.

As I said in my posting, I have no illusions that PDA is the perfect solution to every problem in libraries, or even a perfect solution to the fundamental problems of real-time access for researchers. I believe only that it represents a radically better solution, and one that we should be investigating both energetically and with rigorous care.

Is there anyone here with experience and knowledge of how UK/European libraries acquire needed STM material via the scholarly book trade system? If yes, might they describe if that system is in any way closer to PDA, or not?

When discussing the immediate and long-term value and implications of PDA, it is important to consider the resources it liberates, resources (mainly money and time) that institutions can then apply to parallel and other areas of need. Rick is correct in writing that PDA will not help a library spend less money, but it can help a library spend money more effectively, giving limited budgets more mobility. In this sense, PDA becomes part of a larger strategy to address coverage.

And, to switch gears a bit…it strikes me that generally PDA discussions and literature lack data, or more specifically, a review of the available data. At the moment, I think this hole is especially detrimental to publishers. A pessimistic air so often frames the conversation about what PDA will do to publishers and the scholarly output. I don’t trust this. PDA may push already dying business models to the point of extinction. But, with that comes opportunity and the data points to lots of it (at least from what we’re seeing at EBL). To offer a couple examples, at EBL, unique title coverage and usage across our PDA library customers is over two times that of library’s building collections via “just-in-case” purchasing. Unique publisher coverage is 1.5 greater. Moreover, revenue is greater and more diverse. I find this data encouraging – both for the academy and the institutions that support it.

Do your data show how long after publication these PDA purchases are taking place? One of the concerns for publishers is cash flow. The advantage of the approval-plan model was that presses got the bulk of their money at the point of publication. PDA has the potential to stretch out that ordering cycle over a much longer period of time, even if the absolute numbers sold at the end turn out to be the same, or somewhat greater.

Good question, Sandy. I don’t have this data on hand, but I should be able to analyze this; I will follow-up.

That said, if true, do you believe that this would make the model unsustainable for publishers?

In this fast-changing environment, with so many variables at play, it is difficult to make any predictions with a great deal of confidence. I’d be more worried, frankly, if PDA resulted in overall lower sales of university press monographs, but we’ll have to see what uptake there is on the new subscription-based ebook aggregations from Project Muse, JSTOR, et al. Possibly that can offset erosion of print sales caused by PDA and other factors.

Hi, Robin —

I think you’re right to point out that PDA offers opportunities for publishers as well as threats, and I think the opportunities are especially rich for relative newcomers to market (like EBL, as compared to, say, a 100-year-old scientific journal publisher). They also may be richer for aggregators than for individual publishers. Most threatened, I think, will be those publishers whose organizations and practices have developed over time around a business model based on print-run publishing and sales revenues powered by large-scale speculative purchasing. To the degree that any company depends for its revenue on models that no longer obtain, that company is going to experience serious disruption if those models are pushed out. To the degree that the company is able to retool in response, there will be opportunities as well. That ability (or willingness) will obviously vary greatly from company to company.

I want to say that I understand Rick’s position. However, PDA is not a sustainable system for most Research Libraries. It could provide patrons what they need and save money at the same time. What is missing in this discussion is the issue of the necessity of a comprehensive collection in order to satisfy accreditation requirements in some disciplines. Much as I love to collect only through PDA, I wary to rely only on it. A mixture of traditional collection development and PDA will be a better model.

Hi, Ibironke —

In the majority of libraries, I agree that a blend of approaches would indeed be the only way to go. I’m not saying that all libraries should go PDA-only; I’m only saying that most libraries should explore PDA and then implement it to the degree that it makes sense for their situations. In some libraries, it will make sense for PDA to be a more central strategy, and in others it should probably be more on the margins.

Leave a Comment