This paper was originally presented at an Association of Research Libraries workshop, "The Specialized Scholarly Monograph in Crisis, or How Can I Get Tenure If You Won't Publish My Book?" at a session September 11, 1997 called "Economics of the Specialized Monograph." It is reprinted here with permission.

Marlie Wasserman has shown the economic predicament of the monograph in its traditional codex form: black ink on white pages bound into a convenient bundle. My task is to consider the new means of presentation and delivery that technology has made possible and to examine the many claims that they will solve the economic problem of the monograph. To give a foretaste of where I will come out, I had thought of naming this talk "Puncturing the Panaceas," or "Debunking the Dreams" or "Eviscerating the Visions."

We all know what some of those visions are: Put it on the Internet, shift bytes not atoms, print on demand, print locally not globally etc. They are all ideas for changing the mechanism by which the text is delivered to the reader. I am not going to consider the different kinds of text that new technologies make possible, but concentrate on the traditional linear text that still seems the dominant mode for expressing academic work in the humanities and social sciences. My focus is on changes in the means by which the text is brought to the reader.

System-Thinking and Faculty Time

When we consider alternatives to the present mechanisms for disseminating the work of scholars, we cannot look at publishing in isolation. The academy is a complex life form in which the transfer of ideas in written documents is one of the vital fluids. Changing that apparatus of circulation has many other impacts on the organism. This is why the threats to the survival of the monograph are not merely a matter for academic publishers and librarians, but a central issue for faculty members, university administrators, and funding agencies. A change in one part of the system has impacts throughout the system, and those impacts need to be considered in evaluating the merits of proposed changes.

For example, the universities cut their library budgets, the libraries consequently cut their book purchases, and so demand for books falls. More academic manuscripts become uneconomic for publishers and are declined. The scholars, eager to get their work disseminated, turn to self publication or departmental publication. Those routes do not eliminate the costs of publication, they merely shift them and very often hide them in other university budgets. In fact they probably increase the total costs, as it is unlikely actually that the work will be done either as well or as cheaply as it would be by professional-publishing people — the economic benefits of specialization are well-attested in every other industry, after all. But what is worse is that people who would otherwise be doing research and teaching, or who would otherwise be support staff enabling those crucial activities, are now engaged in the work of publishing to the detriment of the time available for teaching and research.

There are further cost increases because those alternative-publication routes do not provide the reliable quality indicator that a publisher's imprint provides. The scholar-reader will not be able to focus her scarce reading time on work known to be of high quality, but will spend substantial reading time on material that turns out to be of low quality — another erosion of scholars' research and teaching time. And, harder to quantify, informal-publication methods make it much more likely that valuable works will not reach their full readership and that good work will not be fully incorporated into the corpus of scholarship.

Finally we come to the sensitive issues of hiring, tenure and promotion. Without prestigious imprints to use as markers of the distinction of a scholar's work, committees are going to have to spend much more time judging the scholar's work. The already-onerous task of evaluating a scholar will be substantially increased — at every step in her career.

"The academy is a complex life form in which the transfer of ideas in written documents is one of the vital fluids"

Thus a modest change on one area has effects throughout the system. And savings achieved by one element in that system are eroded, possibly wiped out, by extra costs elsewhere in the system. As Peter Senge author of The Fifth Discipline[1] observed: "in many systems doing the obvious thing does not produce the obvious desired outcome." Each participant does the obvious and rational thing, but the systemwide result is unforeseen and perverse. So, as we look at alternatives to the current system, it is important to try to think, not just of the immediate cost, but of the system-wide consequences. And I would say to senior administrators that the lack of mechanisms to effect this in universities is a crippling disadvantage in a rapidly changing environment.

As should already be clear, any systemwide evaluation must take account of the impact on the time of faculty members. Faculty time, both as teachers and researchers, is the most valuable resource in the university. Effective use of that time is crucial to the success of the university in fulfilling both its teaching and research missions. Thus time diverted from those activities to prepare and publish their own manuscripts, time spent to search for materials that are no longer available through well-established channels, time spent reading things that prove after an hour or two to be valueless, increased time spent evaluating their colleagues for tenure — these are all costs, true costs, using the scarcest resources of the academy.

Another example of failing to consider systemwide costs is the common practice of university presses to request that the author provide camera-ready copy. That eliminates our substantial typesetting costs, but it takes a large amount of the scholar's time. The costs have not been reduced, they have been shifted. Indeed, as the work is now done not by a medium-skilled typesetter but by a highly-trained, highly-paid academic, the system-wide costs have undoubtedly been increased. However we will continue to make that request because the university presses, to meet stringent financial criteria, are forced to make narrow, not systemwide, decisions to save costs.

Effect of the Delivery Mechanism on the Publishing Process

I will now explore how changes in the delivery mechanism may affect the publishing process. The conventional publishing process can be divided into three components. The first is file preparation. This involves evaluation and selection, copyediting, typesetting, and proofreading, and ends with transfer of a file to the printer. The second is the delivery mechanism, and involves printing, binding, shipping, and warehousing. The third, not so neatly separable chronologically, involves marketing, order-taking, customer relations, money collection, and royalty payments.

This third component can be set aside for the purposes of this talk. Whatever the delivery method used, we will have to market, we will have to charge customers, collect money and pay royalties. Those activities are continually changing as technology advances but, for example, we are using the Net to advertise books and we would do that also for digital products. Our choice of delivery mechanism will not make much difference to those functions.

What we do need to consider is whether the file preparation costs are affected by the delivery mechanism or whether our exploration of alternatives can take those file preparation costs as given and concentrate only on the costs of the delivery mechanisms themselves.

If there is value in the paper world in careful selection, in guiding author revisions, in copyediting and proofreading, that value carries directly over to alternative delivery mechanisms. Typesetting — the other element in the file preparation process — needs a little more discussion. Typesetting today is actually a matter of coding: A person works through the file containing the text, inserting codes to ensure that the material displays correctly whether the medium of presentation is a sheet of paper or a computer screen. For a given level of functionality the density of coding is about the same, and increasingly the coding is just the same. So, although the delivery medium still affects the coding to some degree, the costs are not going to be significantly different.

Note that I did say "for a given level of functionality." If links and multimedia features are incorporated, the coding costs of the digital medium are likely to be much higher. I am endeavoring in this exercise to restrict my attention to something that could appear in either form, comparing like with like.

In what follows, I will therefore assume that the file-preparation costs apply equally in all delivery media. It is thus the costs of the different delivery mechanisms that can make differences to the economics of publishing the scholarly monograph. At The University of Michigan Press I estimate our delivery mechanism costs to be about 25 percent of total costs; looking at Marlie's figures, I get the sense the percentage at Rutgers University Press would be a little lower. So even reducing delivery-mechanism costs to zero would not bring radical cost savings. Nonetheless even modest savings would be most welcome.


There are three general categories of solution: delivery of a codex but by nontraditional means, e.g. Docutech; delivery of the text on a tangible but nontraditional medium, e.g. CD; and delivery of the text in intangible form, e.g. the Internet.

"Actually shipping atoms is not expensive"

Delivery of a codex but by a nontraditional system.

A Docutech has lower set-up costs (although other printing machines are rapidly catching up) and requires less in the way of support systems and skilled operators.

The first characteristic means that costs rise much less dramatically as quantity decreases than with conventional machines. Unfortunately, their costs per copy are much higher than for conventional machines at even moderate quantities. So even for the small numbers we are now able to print, the Docutech would be substantially more expensive. Of course we still have all the file-preparation costs. So as the number of copies decreases, each copy must still support a larger and larger share of those file-preparation costs. Docutechs are more expensive for print runs in the hundreds and are thus not a solution for the first printing of a monograph.

At very small quantities they are cheaper and thus one way they are being used is to permit very small additional printings to keep books in print. A reprint of course does not have to carry all the file preparation costs, which should have been covered by sales of the first printing. Such short-run Docutech reprints, because of the costs, only make sense for books priced at well over $30, but it does mean some monographs with continuing modest demand can be kept in print.

The second characteristic — less support systems needed — does suggest that distributed printing might be economically feasible with the file distributed electronically to a Docutech close to the customer and printed there. That Docutechs do not have the capability to bind adequately could probably be solved. However the disappointing truth is that the cost discrepancy between a Docutech and conventional printing machine is substantially greater than the cost of shipping the book from a central point to the customer. Despite Nicholas Negroponte's frequent arguments for sending bytes rather than shipping atoms, actually shipping atoms is not expensive and when time is not of the essence, central production and worldwide distribution of the book still offers the lowest costs.

However, taking a narrow viewpoint, this approach might permit the publisher to shed the costs of producing the final copy, leaving that to the library or individual purchaser. System costs would increase as the books would be more costly than under conventional arrangements and the books would be less durable and less well printed. But the university presses would have shifted some of their costs onto other parts of the system and reduced the immediate problem for them of the economics of the monograph.

That seductive but unconstructive path aside, I do not see any general solution to the economic problems of the monograph in the various ideas for transformed production of the traditional codex.

Delivery of a tangible item that is not a codex

Here I am envisioning delivery of a conventional text on some form of computer-readable medium — a CD being the most likely choice.

We still have file-preparation costs. But we now face the risk of additional costs if we take advantage of the extra functionalities that the CD offers. Succumbing to that temptation would mean additional coding costs, design costs, software license fees, and testing. Even for a simple product there will be some minimal additional testing needed that the codex does not require.

Restricting ourselves to a simple monograph on CD, none of the file-preparation costs of the codex can be discarded. The actual costs of preparing the physical CD can be summarized:

Mastering $600
Replicating + Jewel Case $1 per copy
Printed matter $3 per copy (4-color)
Box $1 per copy

Thus 600 copies, a typical run for a monograph in codex form, would cost between $3 and $6 per copy — in roughly the same range as the equivalent costs for a book (Marlie's figures showed a unit cost for a hardcover of $5.00). However if we were to shift completely over to this delivery medium, we would also see a significant reduction on our required warehouse space. Halving our warehouse space (an optimistic assumption, I suspect) would reduce our total costs by around 1 percent.

Thus a shift in to this delivery medium would bring at best some very modest savings. It looks to me as though 10 percent would be a pretty optimistic estimate.

The distribution system poses problems: Bookstores are decreasingly eager to carry CD-ROMs and libraries are certainly unenthusiastic about handling them. So there are real problems in the chain that links publisher to reader. It seems on present experience unrealistic to expect as many CDs to be sold as we would have sold of the codex.

"Because it is possible, there is a considerable temptation to add bells and whistles"

Now consider the systemwide implications: The overwhelming evidence is that people do not want to read books on screen. The evidence also is that people read quite a bit more slowly on screen than on paper [2]. So if the monograph is delivered on CD, it is easy to foresee that most readers will print their own copies off a relatively expensive laser printer. And whereas the single copy of the book in the library serves numerous readers, with delivery on CD each reader is likely to print off his own copy. We are thus likely to see many more impressions in total made by a method that is more expensive per impression than a standard book printing machine. Of course, those printing costs will not be the publishers', so we could happily go ahead down this path, but the costs will hit the universities, albeit dispersed over many separate accounts.

We cannot ignore questions of the life of CDs and the likely need regularly to migrate the files to new computers, new operating systems and, possibly, new text-encoding systems. Those issues threaten significant costs, although probably not to the publisher, but definitely to the system as a whole.

So I predict that widespread use of CDs to distribute monographs might bring a small reduction in the publisher's costs. However the benefits of that might well be overwhelmed by the fall in the number copies sold and thus in revenue. And I do expect that systemwide costs would increase.

As a coda to this discussion, I need to mention the possibility of delivering multiple books on a single CD. While file preparation costs would not be changed, production and distribution costs per book could be very small. Major libraries aside, how often would there be customers who wanted and would pay several hundred dollars for a defined set of (say) 40 books? We could produce a unique selection for each customer, but my hunch is that would impose a nontrivial amount of new overhead in taking orders and manufacturing. I am sceptical so I will not spend more time on this.

The intangible product

Delivering a monograph over the Internet, just as with a CD, does not change the file-preparation costs. Again as with a CD, because it is possible, there is a considerable temptation to add bells and whistles that increase those upstream costs considerably. Indeed it would be much harder to present a text over the Web without adding a rich offering of internal and external links. So there is a strong force leading to increased file-preparation costs.

By contrast with the comparable (or perhaps higher) file-preparation costs, the costs per copy of this delivery mechanism are extremely low. While there are the new costs involved in running a server, shared over many books the cost per monograph should be reducible to a minuscule figure.

Thus I would be hopeful with such a system in place that the costs of publishing monographs could be reduced by between 20 and 25 percent.

We do have to assume that there will be mechanisms for controlling access to the file and collecting money for that access. If many get free access, then the problems in the paper world consequent upon small sales and inadequate revenue would just be magnified in the networked world.

We must again look at the systemwide impact. As before, it is likely that most readers will print out the text and so the cost arguments made with respect to the CD apply equally here: Much if not all the savings from having no tangible product will be eaten up in the costs of making it tangible by inefficient methods. Again as with the CD, we do have the long-term problem of ensuring the files remain accessible as computer systems and coding standards change, and as storage media deteriorate.

There are more major system impacts that need to be looked at in such a transition. Much has been said about the problem of tenure with an electronic publication. However, once there is some clear sense that a respected imprint on paper can maintain the same standards of selectivity in the networked environment, I believe tenure committees will allow electronic publications comparable weight to that of paper publications. Of course, I guarantee you each member of the committee will print out a copy of the networked text.


I hope I have succeeded in puncturing the panaceas. I am sure that none of the alternatives to the traditional system can offer the kind of major cost savings that would make the fundamental problem of the monograph go away. The most prevalent error that has led people to believe in those panaceas is to ignore the value of scholars' time and forget that it should be directed not to mundane publishing chores but to the teaching and research for which faculty are uniquely qualified, expensively trained and rigorously selected.

But if there are no panaceas, there are opportunities that do offer some easing of the current economic predicament. However, those options are mainly cost shifting rather than systemwide cost reducing. Where there are opportunities to take advantage of new technologies, the implied system changes are not in the least trivial — they are major.

And finally, we publishers are driven in part by the wishes of the authors we woo and in part by the readers we serve. There is no pressure from either group for anything other than the traditional codex, carefully edited, nicely produced, and energetically marketed. A shift in delivery mechanism would not be popular and would require much preparation and education of the relevant communities.

Colin Day is graduate of Oxford. After gaining a Ph.D. in economics at Stirling University in Scotland, he spent several years as a Lecturer in Economics before joining Cambridge University Press as Economics Editor. Transferred to Cambridge's New York office in 1978, he later became Editorial Director there. One of his achievements during that period was to launch an electronic journal in 1985. In 1988 he became Director of the University of Michigan Press. During his term as President of the Association of American University Presses, he became active in the policy debates over copyright and electronic publishing. He was a member of the AAU Task Force on Intellectual Property Rights and is now a member of the CIC Copyright and Intellectual Property Committee.


1. Peter Senge, The Fifth Discipline, Doubleday, 1990.return to text

2. Only anecdotal, I would love to see some real research results on this.return to text