The Association of Learned and Professional Society Publishers (ALPSP) is the international trade association for society and other not-for-profit publishers. Such publishers have a particular concern to serve the needs of their scholarly communities and thus to publish in as 'scholarship-friendly' way as they can afford to. There are two dimensions to what they can afford: not only do they need to cover their costs - including overheads and reinvestment in the publishing business -, but also as societies they may be very depending on the proceeds of publishing in order to subsidise other activities such as reduced membership fees, conferences or scholarships.
What constitutes 'Scholarship-Friendly Publishing'? To answer this question, it is necessary to find out what the majority of scholars actually want, rather than only to listen to the vocal minority. ALPSP has therefore carried out two major studies in which questionnaires were sent to journal contributors to find out what was important to them in the publishing process.
In the first study, What Authors Want (Swan & Brown, 1999), which was carried out in 1998/9, questionnaires were sent - courtesy of our members and other helpful publishers - to 10,970 journal contributors, carefully selected to give a balance between subject areas and geographical location. We received 3,218 replies (29.3%). This survey focused on what authors wanted out of the publishing process in general. The second study, Authors and Electronic Publishing (Swan & Brown, 2002) was carried out in 2002. This time 14,643 questionnaires were sent out, carefully balanced as before; however, only 1,246 replies (8.5%) were received, possibly because we relied on completion of an online form rather than sending a paper version to each person. The focus in the second survey was on respondent's views - both as authors and as readers - to both traditional and new forms of electronic research communication.
Authors were very clear about their most important publishing objectives. 33% put 'communication with peers' first, and 'career advancement' came close behind with 22%. Other objectives were much less important: 'personal prestige' (8%), 'funding' (7%) and 'financial reward' (1%). They were also asked how these key objectives were achieved. The most important method was by 'communication with the widest possible audience'. Joint second were 'publication in high-impact journals' and 'the quality of peer review'. Next came 'retrievability through abstracting and indexing services', then 'speed of publication' and finally 'enhancement of personal publications list'. Thus authors clearly seek a balance between wide visibility of their work, and publication in prestigious journals.
In the second survey, we also asked respondents, both in the role of author and in that of reader, to evaluate the importance of various features of the traditional journal. The results were interesting (and encouraging for publishers):
|Gathering articles together||71%||54%|
|Selection of relevant, quality- controlled content||60%||39%|
|Checking citations/adding links||46%||28%|
This suggests that the notion of the journal as an convenient 'envelope' for pre-selected articles - selected by the editors for relevance and interest, as well as for quality - is likely to remain important to both authors and readers; however, some of the other things publishers do, which we believe add value, are perhaps not as highly rated as we would like to think and if cost reduction became paramount, it might be possible to streamline or even omit them altogether.
We were particularly interested to find out authors' and readers' views of the self-archiving movement (the Open Archives Initiative itself, of course, is not limited to freely available material, but personal, subject and institutional archives do in fact make all their content freely available). Interestingly, although a growing number of publishers do now permit self-archiving of pre- and/or post-publication versions of authors' articles, this does not seem to have damaged subscription levels yet. However, if we ever reached the stage where the majority of papers were archived in this way, and where all the archives were readily cross-searchable through the consistent use of metadata and metadata harvesting, then it seems inevitable that the viability of the very journals on which they draw would be seriously threatened.
We made the distinction in the survey between 'preprints' (pre-publication versions of papers) and 'reprints' (post-publication). 32% of authors considered preprint archives important, but only 11% actually deposited their own papers in them. As readers, 78% did not look at preprint archives at all - and in fact, many had never heard of them; familiarity was, understandably, high in high-energy physics but remarkably low in many other fields. It may be that some subject areas are more receptive to self-archiving for a number of reasons:
|·||They are familiar with the 'preprint culture' in the paper world|
|·||The group is fairly small, so that they are familiar with other researchers in their area|
|·||Research is expensive (it uses very expensive equipment, for example) so that, in effect, it is peer-reviewed before it is even done|
|·||The subject area is 'pure' rather than 'applied', so that authors and readers are more or less the same group|
|·||There is no risk to life and health through the release of unreviewed content (as could be the case in medicine)|
The figures were higher for reprint archives - 62% considered them important (though some of the answers suggested that published collections, such as Elsevier's ScienceDirect, were considered as reprint archives). But again, only 11% of authors actively deposited their papers in them.
The Open Access model (where journals are free to readers, and the costs of publication are covered by an up-front fee which is paid by the author, their institution, or - most probably - their research funds) has come to prominence since these surveys were conducted. However, it does seem to have considerable attractions as compared with self-archiving and organised retrieval. if the numbers work, it should make it possible to preserve what is most valuable about journals, while at the same time giving free access to all. It also has the attraction that it would keep in step with growth in research output - which, of course, library acquisition budgets have not been able to do for many years now. However, it is not yet clear whether it will work in all disciplines; the costs which need to be recovered may be more than authors in some disciplines can pay if their research costs and thus grants are typically low (for example, in the humanities). Even in those subjects where research budgets might be adequate to cover publication costs, the full cost of all the functions currently carried out (which many publishers estimate at between $2000 and $3000) may be unacceptably high, so some functions may have to be streamlined or eliminated in order to arrive at an acceptable level of fee. Even then, publishers generally seem to accept that the returns from their publishing activities may be significantly lower than at present.
One of the problems in establishing a reasonable level of fee is that, for a long time, authors and readers have been unaware of much of what the publisher actually does - if asked, they might tend to suggest 'peer review' and nothing else. However, as publishers know (but have perhaps been bad at explaining to others), there is a great deal more to it than that. Publishers create new journals, carrying out market research and then investing money in the new publication - they very rarely start even covering their costs until 5-7 years after launch, and this period is steadily getting longer (and, of course, some new journals never make it at all and are closed down leaving a substantial deficit). Then they manage the process of peer review (even though the review itself is carried out by - generally unpaid - academic experts); they work with their journal editors (who are usually paid, though modestly) to select and collect together the content which the readers will find most useful; they edit and improve the content, the language and the presentation. In addition, they exercise 'quantity control' as well as 'quality control', making sure in most cases that the journal does not get so large that it becomes less useful to the reader and, if necessary, increasing the rejection rate (and thus the quality) to achieve this. Very importantly, they make the journal and its contents visible - marketing in its truest sense - by making sure that it is included in all the right databases and that all the necessary links to and from the articles are in place. And they ensure that the content is preserved for the future, whether by the publisher itself or - more likely - by national and other libraries.
Obviously, all of these processes cost money. Costs are not, in fact, coming down as much as publishers or their customers hoped with the introduction of electronic journals. Print, paper and carriage costs do disappear (though not until customers are happy to accept e-only); however, for most scholarly journals, with very low print-runs, these only represent a minority of the costs - around 80% of the costs are incurred pre-publication and do not change with e-publication (Tenopir & King, 2000). In addition, e-publishing brings new costs - continuous development and updating of the computer system, 24/7 customer support, negotiation and sale of licences, and maintenance of an access control system. Only the last two - licensing and access control - would disappear for an open access journal (and unless a publisher's journals were all entirely open access, this saving could not be made).
What is more, publishers do not just need to cover their direct publishing costs. They also have overheads (staff, equipment and premises costs) and they need to be able to reinvest in the publishing operation - to create new journals, develop new systems, etc. In addition, most publishing operations need to make some extra, whether this is called profit (by commercial publishers) or surplus (by not-for-profit ones). It's clear that a commercial publisher is expected to make a good return for its shareholders; it may be less obvious that many learned societies also depend on their publishing returns to fund, at least in part, the other things which they do on behalf of their communities.
Essentially, there are only three ways a publisher can recover those costs - from the consumer (i.e. the subscriber, whether library or individual), from the producer (i.e. the author, his/her institution or his/her research founder) or from a sponsor (such as an institution or other benefactor).
Publishers are, in fact, doing a great deal in response to the demands of scholars and libraries. Much more content is being made available for not very much more money; there is both more breadth (through bundled collections) and more depth (through the creation of extensive back-file archives. The extent to which previously unsubscribed material is being used is impressive - both OhioLink and Academic Press have reported that more than 50% of material used was previously unsubscribed. And the demand for backfiles is also remarkable - the British Library report that over half of the demand for articles from their Document Supply Centre is for articles more than 5 years old.
Publishers are also offering access to many more people. Consortia licences offer access for all the members of a consortium; special arrangements for less developed countries, whether through formal schemes such as HINARI or eIFL.net or directly by individual publishers, provide free or very inexpensive access to a large number of people who previously had none at all. And many publishers are also making their backfiles widely available, either for a charge or - increasingly - free. These arrangements are often not limited to online access (which may not be affordable or convenient), but offer CD-rom or local printing instead - in many cases there is also support for local publishing initiatives.
However, small publishers are necessarily disadvantaged when it comes to putting together attractive bundles of this kind. A number of collaborative efforts have been created to address this problem - for example, BioOne, Project Muse, the American Institute of Physics (AIP) and most recently the ALPSP Learned Journals Collection (ALJC) is a collection of 247 journals from 25 international publishers, which is sold as a single collection on the publishers' behalf by Swets Blackwell. There is a standard publisher agreement (granting Power of Attorney to Swets Blackwell so that they can sign licences on the publishers' behalf), a standard licence (based on the one at Licensingmodels.com) and a standard pricing and revenue sharing model. Customers can also buy three separate discipline-based subsets of the collection if they prefer.
Publishers are also responding to customer demand and becoming more liberal both in their agreements with authors and in their licences with institutions. Many agreements now permit the things which are most often asked for - the ability to post preprints or electronic reprints on personal, institutional or subject websites; and the ability to re-use material in the author's own institution. The increasing prevalence of model licences, often developed by publishers and libraries in collaboration, is also leading to increasingly relaxed licensing terms from many publishers. Publishers are also willingly experimenting with new business models which might better suit customers' needs - not only pay-per-view and online-only, but also the open access model. Publishers can see that the existing subscription model can't be expected to last indefinitely, with the supply of research articles continually outstripping the growth in the funds available to buy them; they all want to find viable ways of continuing to offer what is valuable about their work.
In the first part of 2003, ALPSP carried out another research study (Cox & Cox, 2003), this time to find out what publishers were actually doing about these various crucial issues - there seemed to be no objective source of factual information, so we plan to repeat the study every few years to identify trends. 275 international journal publishers were surveyed, and 66% responded, including all the major publishers. 75% of the respondents' journals were online (higher in STM, slightly lower in Humanities and Social Science). Large publishers, are perhaps unsurprisingly, most advanced in linking to and from abstracting and indexing databases and their own articles; the majority use CrossRef to do so. Smaller, and particularly not-for-profit, publishers are further behind on linking. Pricing models are still extremely variable, but most still use the print subscription as their starting point - this will undoubtedly start to change over the next few years. Most publishers offer bundles either of all their journals or (typically with specialist society and other publishers) in specific subjects, and consortia licences - however, as 40% of the respondents published 5 or fewer titles and only 8% published 100 or more, most of these bundles must be quite small. About one-third of the publishers currently offer some kind of special arrangement for less developed countries. 85% had back volumes available online, mostly dating back to 1997 or 1998 - which suggests that the files are by-products of electronic publication; however, over 20% have earlier back files, and in a number of cases they are engaged in major retrodigitisation projects of complete journals from Volume 1, number 1 - these, of course, cost money and are rarely if ever made available for free. 60% provide continuing access to e-journals if a subscription is cancelled (this is more common among commercial publishers). An alternative approach, adopted by 9% (but very prevalent among smaller not-for-profit publishers) was to make the archive freely available to all after a certain period. About half of the publishers have some kind of formal arrangement for long-term preservation - either working with national or university libraries, or through organisations such as LOCKSS and JSTOR, or on their own.
As far as agreements are concerned, about 60% (particularly larger publishers) allow use for course packs; 50% allow e-reserve use, 40% paper interlibrary loan and 15% electronic interlibrary loan. 17% do not require authors to transfer copyright, and a further 9% are willing to offer a licence as an alternative if the author is unwilling or unable to transfer copyright. Just under half do permit the posting of published articles to web sites, and rather fewer - about one-third, mainly large publishers - also allow posting prior to publication. Over 80% allow re-use of articles in the author's own institution, and 45% allow re-use within her own publications without further permission.
There are a number of other initiatives - many of them in collaboration with the academic and library communities -, which give an optimistic picture of publishers' efforts to understand and satisfy the needs of scholars. ALPSP has produced a Model Grant of Licence which may be copied by publishers who no longer feel it necessary to insist on copyright transfer. The Zwolle Group - a collaborative group of publishers, academics and librarians - is studying copyright management policies. Project ROMEO has surveyed the published policies of publishers on self-archiving. And ALPSP is working closely with the Open Society Institute and SPARC to investigate the Open Access publishing model.
I think it is abundantly clear that publishers are making every effort to be 'scholarship-friendly'. They are taking the trouble to find out what authors and readers, and librarians, need, and their publication practices are becoming steadily more responsive to these needs. What is also clear is that what publishers do is actually valued by both authors and readers; somebody has to pay for this, and although alternative models may help, they need to be explored carefully to make sure that the numbers still add up.
LIBER Quarterly, Volume 14 (2004), No. 1