Roger Clarke's Web-Site

© Xamax Consultancy Pty Ltd,  1995-2024
Photo of Roger Clarke

Roger Clarke's 'Journal-Publishing Cost-Profiles'

The Cost-Profiles of Alternative Approaches to Journal-Publishing

Roger Clarke **

Pre-Print Version of 10 November 2007

Published in First Monday 12, 12 (December 2007), at

© Xamax Consultancy Pty Ltd, 2005-07

Available under an AEShareNet Free
for Education licence or a Creative Commons 'Some
Rights Reserved' licence.

This document is at


The digital era is having substantial impacts on journal publishing. In order to assist in analysing these impacts, a model is developed of the costs incurred in operating a refereed journal. Published information and estimates are used to apply the model to a computation of the total costs and per-article costs of various forms of journal-publishing. Particular attention is paid to the differences between print and electronic forms of journals, to the various forms of open access, and to the differences between not-for-profit and for-profit publishing undertakings.

Insight is provided into why for-profit publishing is considerably more expensive than equivalent activities undertaken by unincorporated mutuals and not-for-profit associations. Conclusions are drawn concerning the current debates among conventional approaches and the various open alternatives.


1. Introduction

The Internet and its associated technologies have unleashed substantial change in many aspects of life, and of business. Electronic publishing has created new possibilities, and in the process called into question many long-established practices. Of particular significance for refereed journals has been the movement commonly referred to as 'open access'.

The impacts of electronic publishing and the open access movement are multi-facetted, and need to be considered from a variety of perspectives. This paper addresses one particular aspect: the cost-profiles of journal-publishers. It does so from the perspective of those who fund the publishing of academic journals, through the pragmatic lens of the systems analyst and management accountant.

A model of journal-publishing is developed that is sufficiently detailed to enable the cost-profiles of various categories of journal-publisher to be assessed. Among other things, this enables yardsticks to be developed for per-article cost, for both print-publishing and e-publishing. The costs are considered largely in isolation from the sources of revenue used to cover those costs.

The paper commences by clarifying the term 'open access'. It then considers the functions that journal-publishers perform, and the categories that they can most usefully be divided into. A generic model of the journal publishing process is proposed. This enables a structured analysis of the costs involved in journal-publishing. The cost-profiles of several mainstream approaches to journal-publishing are examined, and assessed against available evidence. Inferences are drawn.

2. The Concept of Open Access

This paper is motivated by the need for understanding of changes arising from the application of digital techniques to journal publishing. This has given rise to a movement commonly referred to as 'open access' (OA). There is some diversity in the use of that term. This preliminary section examines its flavours, and proposes a framework that accommodates its various interpretations. It reflects a growing literature, in particular Budapest (2002), Suber (2002), Bethesda (2003), Berlin (2003), Suber (2004a, 2004b, 2004c, slides 2-4, 2004d) and Morris (2005, pp. 122-123).

A key motivation for the OA movement has been the removal of cost impediments: "Open access to scientific journal articles means online access without charge to readers or libraries" (Suber 2002). But OA also implies that there are only very limited obstacles to be overcome in order to gain access to the content. The kinds of non-financial 'permission barriers' that open access seeks to avoid are:

From a legal perspective, the term 'open access' implies that the consumer is not constrained by copyright or other laws from making such reproductions as are necessary to enable access in a form convenient to that consumer. From a practical perspective, the implicit assumptions are made that people have access to information infrastructure, including an appropriate device, appropriate software installed on it, and an Internet connection, and have the capability to use that infrastructure.

The concept described above might be usefully described as 'core OA'.

Some commentators would regard any significant compromise to 'core OA' as being unacceptable. For others, however, there are various forms of 'qualified OA'. Examples include:

At the other extremity, some commentators prefer 'extended OA'. In particular, the definition adopted in Berlin (2003) would require that there be no barrier to "distribute" and "transmit" the work. That implies that the copyright-owner must grant to anyone a licence to re-publish the work. Further, it would require that there be no barrier to "distribute derivative works". That implies that the copyright-owner must grant to anyone a licence to adapt it as well. Neither of these is necessary for 'access' as that term is commonly understood.

Both the qualified and the extended interpretations of 'open access' deflect attention away from the key issue, which is the ability of people generally to get access to information generally, in particular reports of the outcomes of research work. In order to focus on that concern, this paper adopts the 'core OA' notion of 'free and unrestricted access' to content without 'price barriers' or 'permission barriers' (Budapest 2002, Suber 2004a).

A further distraction arises because some authors conflate open access with the ePrints / self-archival movement. An ePrint is "the digital text of a peer-reviewed research article, ... before and after refereeing" ( 2005). An unrefereed paper is published as a 'preprint', to enable colleagues to gain access to it (Harnad 2002-). After an accreditation process, and usually considerable further development, a 'refereed article' is published by the journal whose editor and reviewers performed the accreditation. Subsequently, the final version may also be published by the author as a 'postprint'.

Preprints are extraordinarily important to the dissemination of knowledge, and the author has contributed to discussions in the area (in particular Clarke 2005a, 2005c and 2006). The focus of this paper, however, is not on ePrints, preprints, or postprints. It is on refereed articles, with particular reference to the still-new context comprising digital publishing, a very widely available information infrastructure, and demand for open access to reports on research.

The OA movement represents a challenge to the large-scale and highly profitable journal-publishing industry. A major contribution of this paper is the development of a model that throws light on the cost aspects of the current, lively confrontation.

3. Journal Publishing

In order to generate cost-profiles, a framework is required within which the costs involved in journal publishing can be identified and estimated. This section develops such a framework, based on a definition of terms, an assessment of the functions that journal publishers perform, and consideration of the key characteristics of the various kinds of journal publisher.

3.1 The Nature of an Academic Journal

Throughout this paper, the conventions are adopted that 'an academic article' is a written expression of current and extended information about some specific topic within a discipline or research domain, and that 'a journal' is an outlet for academic articles. The term 'journal' may, however, mean many different things to different people in different contexts. In particular, the term may refer to:

This paper contemplates all of these senses of the word.

The new electronic context has created the possibility that journals may be deconstructed, and converted from a 'place' to a 'space'. This could be achieved through a mechanism such as as a certificate, signed by the editor, and associated with the paper (e.g. Smith 1999). Under this vision, the repository function would be outsourced (to authors, institutions and associations), the infrastructure function would depend on mainstream tools, and the revenue-generation role would largely disappear, or be absorbed by the repository operator. The core quality assurance role would remain in place, probably supplemented by a contents-page that lists the accredited papers, and for each of them provides links to a copy/ies in whichever repository/ies it is known to be in.

3.2 Journal Publisher Functions

This paper specifically excludes from its scope the front-end activities of research, authoring, self-review and revision; and the back-end activities of accessing and using published articles. The focus is on the intermediate activities between those research production and research consumption activities. The trigger for journal-publishing activity is the receipt of a 'submission', that is to say a draft paper for review.

Various approaches have been adopted to explaining what, broadly, it is that journal publishers do. For example, Geyer-Schulz et al. (2003, Figure 1) suggest six 'core processes of value-adding activities': content direction, content packaging, market making, transport, delivery support and services, and interface and systems.

In previous research that the author has conducted in this area, including Clarke (2006), the following additional functions of journal publishers have been identified:

One of the few contributions that has adopted an approach similar to this analysis is King & Tenopir (1998). Their cost analysis distinguishes five "publishing components":

King & Tenopir overcame some of the weaknesses of earlier studies, which had failed to reflect quite basic differences between fixed, variable and average costs. The present study differs, however, in that it is undertaken at a deeper level of granularity, with a stronger focus on electronic tools and electronic dissemination than was appropriate almost a decade earlier, and in the context of a maturing OA movement.

A further difference between 1998 and the present is the extent to which journals' primary, large-volume business processes are supported by electronic tools. Commercial publishers have invested large sums of money in proprietary tools intended to deliver or sustain competitive advantage. For example, Elsevier's ScienceDirect was declared to a House of Commons Committee to have cost Stg[[sterling]]200 million (USD$360 million) to develop and to require ongoing investment of "well over Stg[[sterling]]100 million" (Davis 2004).

At the other extreme, not-for-profits have sought out inexpensive alternatives. Most significant among these have been, at the back end, Google search facilities, and at the front end, the Open Journal System (OJS 2006). OJS was released in November 2002 following 18 months of development, and by early 2007 claimed to support in excess of 900 eJournals (Willinsky 2005). First Monday converted to OJS in October 2007.

How resources are mobilised to support journal-publishing is not the primary focus of this paper. Sources of cash, kind and effort include journal-subscribers, sponsors of various kinds, authors and especially members of the community providing gratis effort. In most cases, multiple sources are used. A fuller discussion of business models to support journal-publishing is in a companion paper, Clarke (2005b).

3.3 Journal Publisher Categories

The functions identified in the preceding sub-section are performed by organisations of considerable diversity. Dimensions of difference include the following:

Hovav & Gray (2001) identified the diversity of approach adopted by e-journals in particular. Even within the 'open access' movement, as many as nine variants have been identified (Willinksy 2003b).

There is a strong tendency for the characteristics to cluster, enabling concentration on a few categories. The taxonomy used in this paper is shown in Exhibit 1.

Exhibit 1: A Taxonomy of Journal-Publishers

Term Used
An informal association of a modest number of people with a common interestUnincorporated Mutual
A formally constituted not-for-profit association of individuals, usually within a particular discipline, profession and/or geographical region; or a not-for-profit business unit within such an organisationNot-For-Profit Association
A for-profit corporation, or a profit-oriented business unit of a not-for-profit associationFor-Profit Publisher

3.4 A Cost Framework

A framework is required that will support the identification of cost-elements, and analysis and estimation of costs for the three categories of journal-publisher identified above, under various circumstances.

One possible source of such a framework is a process model of journal-publishing. For example, Geyer-Schulz et al. (2003, Figure 5) identify two categories of business processes, which they refer to as 'traditional activities': initiation of content-reviewing, process of peer-reviewing, abstracting, paper and/or digital print, review and distribution; and 'potential future activities': cataloguing, structuring, personalized data evaluation, stockkeeping, licensing, portal services, lending, support/user guidance, research tools, database services and payment services. Models of this kind are valuable, but they do not satisfy the needs of this paper.

Another alternative is to apply theories of economics to the problem as is done in Page et al. (1997), Friedlander & Bessette (2003) and McCabe & Snyder (2004, 2005). But the kind of framework necessary to support this analysis is less concerned with micro-economic theory, and more with the pragmatics of managing a business.

Management accounting draws a distinction between costs that are readily associable with the outputs of the production process ('direct costs'), and those which are not ('indirect costs'). Applying these concepts in the context of journal publishing, however, is problematical, because the outputs are multifaceted (comprising pages within papers within Issues, and copies of Issues within complete production runs).

In this context, the more useful distinction is between:

The framework used in this paper reflects prior analyses that are evident in a gradually expanding literature, notably Odlyzko (1997), King & Tenopir (1998), Bot et al. (1998), Rowland(2002), Willinsky (2003a), Suber (2004c), King (2004), Hawley (2004) and Morris (2005). It only partly reflects the very substantial differences that exist in the communication, accreditation, debate and consensus styles of different disciplines (Valauskas 1997).

The elements of the framework are identified in Exhibit 2.

Exhibit 2: Cost Model Overview

  • Establishment
  • Operations

    • Submission-Related
    • Article-Related
    • Issue-Related
    • Generic

  • Infrastructure Maintenance
  • Financial Aspects

The framework's scope expressly excludes the activities of the author. In relation to research, writing and re-writing, this is appropriate because this paper's focus is on publishing. It could, however, be argued that some of the efforts of authors should not be excluded, e.g. where they perform publishing-related tasks, such as reference-checking, formatting of tables and illustrations, formatting of text in accordance with the journal's norms and template, and final, pre-publication review of the edited version of the paper.

The framework expressly excludes consumers' costs in accessing the journal, and the costs of access-intermediaries such as cataloguers, abstracting services, post-publication reviewers, third-party repositories, mirror-site operators, consolidators, and research assistants.

4. The Cost-Profile Analysis

This section applies the above framework, in order to identify the items within each element that may give rise to costs. For each element, a brief discussion is provided. The purpose of the discussion is to enable estimates to be developed for each element, under various assumptions about the profile of the publisher and the journal.

This and the following section draw on the author's experience as an author who has interacted with 20-30 journals in multiple disciplines, as an officer of several disciplinary associations, as an editorial board member for various journals, as an editor of a dozen journal special issues and conference proceedings, and as an author, editor and publisher of a dozen monographs. The analysis and estimates have also benefited from conversations with, and informal reviews by, editors and publishers in each of the categories of journal publishing.

In the tables below, the following terms are used:

4.1 Establishment

Exhibit 3 presents the analysis of the first segment of costs, which are those that relate to the formation of the undertaking.

Exhibit 3: Establishment Costs


Conception and articulation of the journal's name, scope, philosophy and modus operandiThis is generally undertaken by senior academics, but may require time from paid executives
Preliminary negotiations within the intellectual communityThis is generally undertaken by senior academics
Preliminary negotiations with potential providers of operational resourcesThis is generally undertaken by senior academics
Preliminary negotiations with potential providers of infrastructureThis is generally undertaken by senior academics, but may require time from paid executives
Acquisition of investment and working capitalThis may be undertaken by senior academics, but may require time from paid executives
Appointment of Board, Editor and Editorial Committee(s)This is generally undertaken by senior academics, although appointment of the Editor may require time from paid executives
Accumulation of Referees ListThis is generally undertaken by senior academics
Acquisition of infrastructureThis may be undertaken by senior academics, but may require time from paid executives. Web-sites may be run using gratis, open source tools (ranging from simple to highly sophisticated) or commercial tools involving expensive licence fees. Willinsky (2003a) reports that the capital cost of the Open Journal System was about $60,000
Acquisition of operational resourcesThis is generally undertaken by senior academics, but may require time from paid executives
Acquisition of intellectual property (logos, trademarks, copyrights, licences)This may be undertaken by senior academics, or require time from paid executives. The intensity of effort invested varies from virtually nothing to substantial
Preparation of formal components of the printed journal and web-siteThis is likely to be undertaken by senior academics, but may require time from paid staff
Preparation of web-siteThis may be undertaken by junior academics or students, or require time from paid staff
Announcement to the communityThis is generally undertaken by senior academics, but may require time from paid executives, and perhaps paid advertisements
Issue of initial calls for papersThis is generally undertaken by the Editorial Committee

4.2 Operations

This section draws together the various aspects of a journal's cost-profile that relate to its ongoing operations. Some costs are associated with the submissions that arrive. Other costs arise in relation to the articles that are accepted for publication. Yet others are associated with the discrete Issues that the journal publishes. Further operating costs, referred to in this paper as 'generic', are independent of each of those groups of activities.

(1) Submission-Related

Exhibit 4 deals with costs arising in relation to submissions of various kinds, including papers, notes, and letters. These may arrive by the score, in many cases by the hundred, but in some cases by the thousand. The business processes involved include internal correspondence within the Editorial Committee, correspondence with the author, and re-submission. Considerable effort and time are invested in what is conventionally referred to as 'peer review'. For each submission, the correspondence generated among referees and the editorial team will generally be an absolute minimum of 6-8 (1 acknowledgement, 2-3 requests, 2-3 reports and 1 notification), and will generally be between 10 and 50 items per submission.

Exhibit 4: Submission-Related Costs 


Receipt, acknowledgement and managementThis is generally undertaken by the Editor, Editorial Committee and Referees, possibly supported by junior academics or students, or perhaps using the time of paid staff
Conduct of the assessment processThis is generally undertaken by the Editor, Editorial Committee and Referees, generally gratis, but possibly with an honorarium for the Editor, and possibly allowances, free advertising or similar partial recompense

(2) Article-Related

The segment of the cost-profile in Exhibit 5 relates to those articles that survive the refereeing process and are accepted for publication. There may be dozens, scores, or in some cases hundreds of accepted articles. For some journals this may be a considerable proportion of all submissions (e.g. 30-60%), but for many it is a small percentage (perhaps 20%, with as low as 10% reported by Nature 2004 and by leading journals in the Information Systems field, such as the Journal of the Association for Information Systems).

Exhibit 5: Article-Related Costs


Production-editingThis is undertaken possibly by the Editor or a member of the Editorial Committee, but more likely by a junior academic or student, or using the time of paid staff. Some of the effort may be outsourced to the author
CataloguingThe preparation, review and formatting of metadata, and entry into appropriate catalogue(s), is generally undertaken by a junior academic or student, or using the time of paid staff. Some of the effort may be outsourced to the author

(3) Issue-Related

A journal is commonly published as a series of Issues. An Issue comprises a discrete set of refereed articles and perhaps other submissions, released at a particular point in time. An Issue of a printed journal typically contains 4-15 articles. The frequency is most typically monthly or quarterly, but may be weekly, bi-monthly, annually, linked to an event such as a conference, or irregular. The periodic cycle of journal issues gives rise to the costs detailed in Exhibit 6.

In the electronic context, the economic incentive to bundle accepted papers into an 'issue' is much less significant, and for some journals the analysis undertaken in this paper may need to treat each article as a separate 'issue'. In addition, markedly different patterns of production and distribution costs arise, depending on whether issues are published in hard-copy, electronic form, or both.

Exhibit 6: Issue-Related Costs


EditorialThis is undertaken by the Editor
Production-editingThis is undertaken possibly the Editor or a member of the Editorial Committee, but more likely a junior academic or student, or using the time of paid staff
ProductionFor hard-copy issues, printing is likely to be either performed by paid staff, or outsourced.
For soft-copy issues, uploading and release of the transmittable format (commonly HTML or PDF), may be undertaken by the Editor or a member of the Editorial Committee, but more likely by a junior academic or student, or using the time of paid staff
ProtectionFor soft-copy issues, protections such as password- and/or cryptography-based locking mechanisms or watermarks may be imposed. Activities like this are likely to be either performed by paid staff, or outsourced, or possibly performed by special-purpose software
DistributionFor hard-copy issues, distribution is likely to be either performed by paid staff, or outsourced.
For soft-copy issues, generation of the issue home-page and issue of an announcement to the subscription-list may be undertaken by the Editor or a member of the Editorial Committee, but more likely by a junior academic or student, or using the time of paid staff

The term 'first copy costs' is sometimes used in the literature, but is of limited use as an analytical tool. One possible interpretation of it is the sum of the submission-related costs - sub-section (1) above (but only for those submissions that appear in a particular Issue), the article-related costs - (2), and some of the issue-related costs addressed in this sub-section (but excluding production and distribution).

(4) Generic

Exhibit 7 brings together those cost-items that arise from ongoing business operations, but that are not directly related to any of submissions, accepted articles or issues.

Exhibit 7: Generic Costs


MarketingThis is a highly variable activity, from effectively nil to substantial.
It may be undertaken by senior academics assisted by junior academics or students, or require time from paid executives and paid staff
Customer relationship managementThis is a highly variable activity, from minimalist records to a substantial database, and minimalist to substantial customisation of services. It includes the costs of collecting revenue.
This may be undertaken by a junior academic or student, or may require time from paid staff
Archive managementThis is a highly variable activity, ranging from leaving the articles in an accessible location, to releasing them from subscriber-only to open access after a period of time, to strongly protected access, including payment facilities for per-view and short-period access.
The more sophisticated facilities may (but may not) require software licences, and time from paid staff
IndexingThis is a variable activity, ranging from leaving the site open to web-crawlers and perhaps pointing to one of them, to using a gratis local search-engine, or a for-fee local search engine, or a sophisticated facility including auto-generated cross-linkages among articles within the journal or a journal collection, or across multiple collections.
The more sophisticated facilities may (but may not) require software licences, and time from paid staff
GovernanceThis involves meetings of the Editorial Committee(s) and periodic reports to stakeholders.
These may be undertaken by senior academics, may possibly require time from paid executives, and may involve travel costs

4.3 Infrastructure Maintenance

Costs may be incurred in sustaining the technical infrastructure and the intellectual infrastructure on which the journal's operation depends. These costs, described in Exhibit 8, are distinguishable from, and additional to, the investment that was necessary to establish them in the first place. They are recurrent, over varying cycles. The convention in investment analysis is to make the simplifying assumption that an annual allowance can be set aside to cover them.

Exhibit 8: Infrastructure Maintenance Costs


Editor and Editorial Committee(s)This is occasional, ongoing activity by the Editorial Committee(s), undertaken by senior academics
A pool of refereesThis is ongoing activity by the Editor and Editorial Committee(s)
Communications channelsThis is generally arranged by senior academics, supported by junior academics or students
Norms for communications and formattingThis is ongoing activity by the Editor and Editorial Committee(s)
Production facilitiesFor hard-copy issues, this is likely to require support by paid staff, or outsourced service providers.
For soft-copy issues, it is more likely to be guided by senior academics, and undertaken by junior academics or students, or using the time of paid staff
Subscription-list facilitiesAs for production facilities above
Distribution mechanismsAs for production facilities above

4.4 Financial Aspects

Exhibit 9 reflects the financial value of capital employed in the endeavour. These costs are addressed differently, depending on the business model employed. Approaches include sponsorship (a 'no-interest loan'), a loan at lower-than-commercial rates, a loan at commercial rates that reflect the assessed risk to the lender, risk/profit-sharing, and fully risk-taking / profit-sharing. Journal undertakings embody risk, but subscription-based journals have also had very considerable up-side potential.

Exhibit 9: Financial Costs


Interest on investment capitalThe assets required to run the operation may be insignificant, or may be gifts, provided by grants, or sponsored.
Otherwise, the financial value tied up in the assets needs to be remunerated as interest payments or dividends
Interest on working capitalAs for investment capital above

5. Sample Cost-Profiles

This section applies the model presented in the previous section, in order to develop an appreciation of the costs involved in various approaches to journal publishing. The first sub-section outlines the approach adopted, and the second summarises the results and provides access to the detailed computations. The final sub-section compares the results against other data evident in the literature.

5.1 Method

The estimates provided below and in the Appendices draw on the author's experiences as enumerated in section 4 above, on the financial statements or budgets of a modest number of unincorporated mutual and association journals, and on interpolation based on the limited information publicly available about the budgets of for-profit journal publishers.

All costs are expressed in $US, and are loosely based on current costs in 'advanced western nations' such as the U.S.A., the U.K., The Netherlands, Canada and Australia. The differences among costs in those countries, probably +/- 10-20%, are of the same order of magnitude as the anticipated error-factors in the estimates. All labour costs are intended to include labour overheads (such as leave and superannuation) and organisational overheads (such as serviced workspace, equipment and utilities).

The model gives rise to an enormous range of possible outputs, depending on the characteristics of the particular journal. To test the model, a small number of mainstream patterns are defined, and the costs computed on an annualised basis. In addition, because of the current topicality of 'author-pays OA', a per-paper equivalent is calculated.

The model expressly avoids imputing financial costs for non-financial sponsorship for journal publishing activities. This is of course highly significant, because the time that academics and other professionals spend on editorial and refereeing activities alone is very large. Similarly, sponsorship-in-kind is provided by employers of academics and professionals in such areas as access to information infrastructure, submission-handling, and in some cases hosting of the electronic publication. It is appropriate that such costs be omitted rather than imputed, because this reflects the realities of managing a journal, and of choosing among alternative journal-publishing models.

A further simplifying assumption has been made. The estimates are for journals whose content is predominantly straightforward text, with supporting tables, and figures that do not require enormous precision. Hence the model does not reflect the additional costs that arise in such circumstances as the following:

5.2 Results

The model was applied to a small set of circumstances. These were selected with three criteria in mind. They were to represent the range of contexts; they were to be realistic, so as to enable meaningful inferences to be drawn; and they were to be sufficiently small in number to be manageable. The results are summarised in Exhibit 10.

Costs for all three categories of journal-publisher distinguished in Exhibit 1 were evaluated. In each case, both hard-copy and eJournal approaches were considered. For the kinds of relatively large-scale journals run by professional associations and commercial publishers, the key features chosen were quarterly issues of 7-8 articles, and 10,000 copies. Journals run by unincorporated mutuals are typically specialised, and the profile evaluated comprised quarterly issues of 5 articles, and 300 copies. The calculations are provided in the Appendices to this paper.

Exhibit 10: The Cost Profiles of Journal-Publisher Categories

CategoryCharacteristicsCostAuthor-Pays Cost
Unincorporated Mutuals   
Unincorporated Mutual hard-copy journal100 submissions p.a., a quarterly issue containing 5 articles, 300 subscribers, with hard-copies produced and distributed by an outsourced service provider$20,000 p.a.$1,000 per article
Unincorporated Mutual gratis eJournal100 submissions p.a., a quarterly issue containing 5 articles, made available on the Web open and gratisEntirely absorbed by sponsorsNil
Not-For-Profit Associations(Assumes publishing is funded from membership subscriptions)  
Association with a single, hard-copy journal10,000 members, a quarterly issue containing 7-8 articles, with hard-copies despatched to all members, and soft-copies available on the Web open and gratis to members, and to all comers after 1 year$112,000 p.a.$3,750 per article
Association with a single eJournal10,000 members, a quarterly issue containing 7-8 articles, with soft-copies available on the Web open and gratis to members, and to all comers after 1 year$22,000 p.a.$730 per article
Multi-Journal Association that publishes 5 journalsSame as above.
Economies of scale and scope roughly balanced against diseconomies
Same as aboveSame as above
For-Profit Publishers (Includes business units within not-for-profit associations that are depended upon by the association to generate a surplus)  
Multi-Journal Commercial Publisher, applying conventional, subscription-based accessMany journals, quarterly issues each containing 7-8 articles, with hard-copies despatched to subscribers, and soft-copies available on the Web to paid subscribers and to others on a per-article basis for a fee. Substantial investment in brand-image, strong competitive stance, aggressive protection of contentHard Copy: $137,000 p.a.
eJournal only: $112,000 p.a.
Hard Copy: $4,600
eJournal only: $3,700
Multi-Journal Commercial Publisher, applying the 'author-pays OA' approachAs above, but funded primarily by authors rather than subscribers, and perhaps without content protectionHard Copy: $127,000 p.a.
eJournal only: $102,000 p.a.
Hard Copy: $4,200
eJournal only: $3,400

5.3 Comparisons with the Literature

Comparison with data in the published literature is not simple, because the authors concerned have made widely varying assumptions, and in some cases it is not entirely clear what they were. This section accordingly identifies key data in the existing literature which is variously consistent with or in apparent conflict with the outcomes of the analysis conducted in this paper.

In Odlyzko (1995) and Odlyzko (1997), the figure of $4,000 per article was suggested for conventional publishing, and $US300-1,000 per article for electronic publishing. These are broadly consistent with the results of the analysis conducted in this paper.

The base configuration used in King & Tenopir (1998) resulted in total costs of $633,745, or $5,125 per article. The differences from the figures reached in this paper's analysis ($127,000 / $137,000 and $3,750 / $4,600 per article) appear to reflect a publication four times the size, all submission-handling performed on a fully-paid basis, higher overheads, and higher printing costs prevalent a decade and more ago.

Of the six journals examined in Hovav & Gray (2001), four were e-journals whose costs were entirely defrayed by sponsorship, mostly from academics and their host-institutions. One was an e-journal publishing c. 80 articles p.a. for an annual cost of $25,000 p.a. This was apparently expended on submissions management and production-editing, for which the model in this paper uses an estimate of $22,000 p.a. The other was a Kluwer print-journal "with electronic presence", costing $200,000 p.a., which is rather higher than the $137,000 arising from the analysis in the present paper.

In Bergstrom (2001), 'first copy costs' were judged to be about $100 per page and the 'marginal subscriber costs' about $.02 per subscriber per page. For the journal configurations considered in this analysis, that results in, for the not-for-profit, an estimate of $42,400 compared with $20,000; and for the for-profit, $210,000 cf. $112,000 or $137,000. Bergstrom was focussed specifically on economics journals, however.

In Rowland (2002), the suggestion was made of $200 per submission for handling costs, which is in line with the $10,000 p.a. figure used in the tables in this paper.

In Willinsky (2003a), pre-production costs ("copyediting and proof-reading") were estimated at $10,000 annually for a typical quarterly journal. This paper uses a similar figure.

In King (2004), author-funded open access publishing was suggested as costing between $500 and $1,750 per article when conducted by unincorporated mutuals and associations, and $3,000-$4,000 per article for commercial publishers. King used a very similar, indicative $100,000 for investment funding "for start-up, capital requirements, future research and development, and operations". He used a fixed figure of $40 per subscriber p.a. for hard-copy production and distribution costs. The model developed in this paper uses the figure of $25 per subscriber p.a., but, unlike King (2004), allows for sensitivity to volume.

In SQW (2004, cited in Morris 2005), estimates are provided of $1,425 to $2,750 per article, "depending on the rejection rate and editorial quality of the journal". The model developed in this paper distinguishes costs per submitted paper and per published paper, but does not treat editorial quality as being a significant determinant of cost. Morris applied a multiplier of 1.6 for overheads and profits, to reach $2,250 to $4,375. These estimates used King's figure of $40 per subscriber p.a. for hard-copy production and distribution. These are not greatly different from the estimates derived from the analysis in this paper.

In evidence to the U.K. House of Commons, Elsevier (2004) stated that "even the highest article fees charged by Open Access publishers today ($1,500) cover only about 40%-60% of the estimated total costs to publish an article of the quality that researchers are used to today". This implies that their own estimate of the average is $2,500-$3,750 per article. This is broadly in line with the inferences drawn from the analysis in this paper.

The BioMed Central's author-pays open-access model (2005) involves costs to each author of a published article of $630-$1,595, depending on the journal. The Public Library of Science (PLOS) began with its rate set at $1,500, but shifted to $1,250-$2,500 in mid-2006. (The variations reflect whether the journal is only available electronically or also in hard-copy, and in the case of PLOS whether the subject-matter is clinical in nature). Butler (2006) has suggested that even the higher figures are unsustainable (but some care is necessary because that report is by a commercial publisher with a vested interest in undermining the credibility of open access). The figures derived from the analysis in this paper suggest that such figures are, on the one hand, very likely to be sustainable for an e-journal, but, on the other, challenging for a print-journal, particularly in the case of high-review-cost clinical subject-matter.

The 'author-pays' movement has spawned 'hybrid journals', which support both author-pays for open articles, and conventional subscription-only arrangements. The first-mover was Springer Open Choice in 2004, which set the price at $3,000 (2005). Blackwell's Online Open model set its price at $2,500 (2005). More than a dozen publishers now offer this option (Suber 2003-). Each journal has to elect to make the option available, and each author has to elect to use it. It is accordingly not directly comparable with any of the cost-profiles developed in this paper. The figures are, however, not greatly different.

In Willinsky (2005), examples are referred to of both an eJournal with zero budget, and "a small group of electronic journals that were spending in the area of $20,000 a year". These are consistent with the analysis undertaken in this paper.

In Willinsky (2006, pp. 227-232), the average of a sample of Canadian social science and humanities journals was computed as 561 subscribers, and costs of either Can$56,620 p.a. (print-based, actual costs in 2003) or Can$24,789 p.a. (eJournal only, computed). The print-based case falls mid-way between two of the models used in the this paper, both in terms of the characteristics and the costs. The eJournal case, on the other hand, is in fairly close accord (USD$22,000 cf. Can$24,789).

Most recently, Willinsky & Ranjini (2007) reports the case of "a small scholarly association with 220 members", which has applied Willinsky's Open Journal System (OJS 2006) to achieve an open access eJournal, with a 'registered alert list' of 450 readers, on a zero budget. This corresponds with the results of the analysis conducted in this paper.

Comparison with the existing literature suggests that the model developed and applied in this paper is broadly in line with prior analyses, but highlights variations that warrant deeper investigation. Importantly, this paper offers something that previous publications lack - an explicit model that enables deeper investigations to be performed.

6. Conclusions

This final section provides some general comments, discusses more specific inferences about cost-profiles, and draws out the implications of the analysis for for-profit journal-publishers.

6.1 General Conclusions

The cost-profile framework in section 3.3, and the cost-profile model in section 4 require examination, criticism and enhancement. The estimates provided for each particular category of journal-publisher in section 5 and the Appendices similarly require study and improvement. The results do appear, however, to be broadly in line with the various estimates available in the literature.

The model has the further advantage of being sufficiently transparent and articulated that it can be applied to analyse the cost-profiles of different variants of journal-publishing, and to conduct comparisons among alternative approaches. The model may therefore offer considerable benefits for many categories of stakeholder. In particular, it may be of assistance to editorial teams, and executives in both incorporated and unincorporated associations and for-profit publishing organisations. It may also be useful for observers of the dynamics of publishing in an era of considerable change.

6.2 Some Inferences About Cost-Profiles

The cost-profiles discussed above are of necessity based on many assumptions, and on data whose quality could reasonably be questioned. With those important qualifications, some inferences can be drawn from the data.

The quantum of sponsorship by senior academics, and by their employers, represents a very substantial proportion of the publishing costs of all refereed journals. As Odlyzko (1997) put it, "the direct costs of a journal article are dwarfed by various indirect costs and subsidies". The sponsorship share appears proportionately smaller in the case of for-profit journal publishing, but mainly because the total cost is higher rather than because the scale of academic sponsorship is significantly lower.

An unincorporated mutual can produce a limited-quantity hard-copy journal for $20,000 p.a., and an eJournal for $8,000 p.a. For costs of this scale, it is not unduly challenging to find sponsors, and to run a gratis eJournal 'on the smell of an oil-rag'. Sustaining the commitment of senior academics would appear to be the most critical survival factor.

Because of the low cost of an eJournal, it can be easily supported by an association with a membership of as little as a few hundred. A hard-copy journal, on the other hand, needs an Association with a few thousand members if it is to 'carry' the journal using only a small proportion of its members' fees.

It is not clear that the net effect of economies of diseconomies of scale and scope is of any great significance, and hence the cost-profiles of publishers of multiple journals may not show much advantage over those of organisations that publish a single journal. Such advantages as do exist may be even less important in the case of eJournals.

For-profit publishers have higher cost-profiles than not-for-profit associations, because of the additional functions that they perform, in particular their much greater investment in branding, customer relationship management and content-protection. The difference is particularly marked in the case of eJournals - a computed per-article cost of $3,400 compared with $730. This point is sufficiently significant that further examination is warranted.

6.3 Implications for For-Profit Journal Publishing

The model invites focus on some important questions about publishing-for-profit in comparison with publishing by not-for-profit associations and unincorporated mutuals. To the extent that the model and the data are fairly indicative of the state of play, it would appear that open access journal-publishing is achievable through not-for-profit channels far more cheaply and efficiently than through for-profit organisations.

The first area of interest is comparisons between cost-profiles. One difference is in hard-copy production and distribution costs. Large associations achieve relatively low per-copy costs because of the scale of their membership lists; whereas many journals published for profit have much smaller print-runs and hence higher per-copy costs. But the most significant reasons why the costs of for-profit publishers are higher appear to be as follows:

The overall impact was recently summarised as follows: "All the evidence shows that non-profit journals are on average ... less expensive ... Studies in various subject areas have consistently shown that commercial journals [levy subscription fees] 2-5 times as much as their not-for-profit equivalents" (Morris 2005). The analysis conducted in this paper suggests that part of the reason for their higher fees is their substantially higher cost-profile. The other part is the super-profits arising from the entrenched monopoly that an established journal represents.

The higher costs of for-profit publishing need to be seen in the context of the benefits that arise from them.

During the long era in which journals were of necessity produced in hard-copy, specialist publishing companies were repositories of expertise on pre-production, production and distribution. The diverse professional and technical skills involved had to be managed, and publishing companies provided operations management expertise as well. There were also potentially significant economies of scale and scope that large operations could convert into cost-advantages.

In the early twenty-first century, pre-production is far less forbidding than it once was. So are production and distribution. And hence so is management. Economies of scale and scope are less significant, and more closely balanced with the diseconomies that arise from having all activities forced into a mould. In the hard-copy world, for-profit publishers continue to offer some benefits, but far less than they once did.

As journals have migrated to dual-mode publishing and to purely electronic formats, the advantages originally offered by for-profit publishers have dissipated. The level of professionalism required to operate an eJournal remains significant, but it is not out of the reach of committed senior academics supported by junior academics and students. Acquisition of infrastructure, and management of infrastucture and processes, are less challenging than was previously the case.

The distinctive differences that remain in for-profit publishing are:

But the primary beneficiaries of these features are the publisher and its owners. Only in the case of for-profit business units within not-for-profit associations are the owners closely associated with an academic community. Academic communities have little incentive to contribute to the funding of sophisticated technical features that are designed to support organistions' strategic and marketing objectives rather than community service. In short, the 'value-add' that for-profit publishers offer appears to be of little or no benefit to academic communities.

For-profit publishers have long been successful intermediaries between the authors and accreditors, on the one hand, and the consumers of refereed articles and their support services, on the other. Since the advent of the public Internet, however, much has been written about the way in which it converts marketplaces to marketspaces, extends the reach of market participants, and creates the scope for disintermediation (e.g. Malone et al. 1987, Brown 2001, Howard 2001). In the new context, are for-profit publishers still needed?

The following possible bases exist whereby for-profit publishers can sustain their role:

The journal-publishing arena appears to be in a state of flux, stimulated by the new economics of Internet-era ePublishing. The framework, model, estimates and findings reported in this paper represent a contribution to understanding how the current dynamics may play out.


  1. An Unincorporated Mutual Hard-Copy Journal
  2. An Unincorporated Mutual Gratis eJournal
  3. An Association with a Single, Hard-Copy Journal
  4. An Association with a Single eJournal
  5. An Association with 5 Journals
  6. A Conventional, Subscription-Based For-Profit Publisher
  7. An Open Access For-Profit Publisher


All items for which URLs are provided were most recently accessed 24 March 2006, unless otherwise noted.

Bergstrom T. (2001) 'Free Labor for Costly Journals?' Journal of Economic Perspectives (Fall 2001), at

Berlin (2003) 'Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities' Conference on Open Access to Knowledge in the Sciences and Humanities, Berlin, 20 - 22 Oct 2003, at

Bethesda (2003) 'Bethesda Statement on Open Access Publishing' Meeting on Open Access Publishing, Chevy Chase, Maryland, April 11, 2003, at

Biomed Central (2005), at

Blackwell Publishing (2005), at

Bot M., Burgemeester J. & Roes H. (1998) 'The Cost of Publishing an Electronic Journal: A general model and a case study' D-Lib Magazine, November 1998, at

Brown D.J. (2001) 'The impact of disintermediation and the new economy on STM electronic information systems' Serials 14 (2001)

Budapest (2002) 'Budapest Open Access Initiative', Open Society Institute, Budapest, February 14, 2002, at

Butler D. (2006) 'Open-access journal hits rocky times' Nature 441, 914 (22 June 2006), at

Clarke R. (2005a) 'A Proposal for an Open Content Licence for Research Paper (Pr)ePrints' First Monday 10, 8 (1 August 2005), at

Clarke R. (2005b) 'Revenue Models for Journal-Publishing in the Open Access Era' Xamax Consultancy Pty Ltd, April 2005, at

Clarke R. (2005c) 'A Standard Copyright Licence for PostPrints', Xamax Consultancy Pty Ltd, September 2005, at

Clarke R. (2006) 'An Open and Closed Case: PrePrints and PostPrints in Digital Repositories' Proc. Symposium on Managing Openness in Digital Repositories, Sydney, 3 February 2006, at

Crow R. & Goldstein H. (2003-) 'Guide to Business Planning for Converting a Subscription-based Journal to Open Access' Open Society Institute, Edition 1.0, January 2003, revised Edition 3, February 2004, at

Davis C. (2004) 'Uncorrected Transcript of Oral Evidence' Enquiry into Scientific Publications, Science and Technology Committee, House of Commons, HC 399-i, Monday 1 March 2004, at

Elsevier (2004) 'Memorandum from Reed Elsevier', Appendix 46 to HCSTC (2004), at (2005) 'What is an Eprint' Self-Archiving FAQ,, at

Friedlander A. & Bessette R.S. (2003) 'The Implications of Information Technology for Scientific Journal Publishing: A Literature Review' National Science Foundation NSF 03-323, June 2003, at

Geyer-Schulz A., Neumann A., Heitmann A. & Stroborn K. (2003) 'Strategic Positioning Options for Scientific Libraries in Markets of Scientific and Technical Information - the Economic Impact of Digitization' J. Digital Infor. 4, 2 (May 2003), at

Harnad S. (2002-) 'Self-Archiving FAQ', at

Hawley J.B. (2004) 'JCI & Open Access' Proc. Conf. Open Access Publishing, Society for Scholarly Publishing, November 2004, slide-set at

HCSTC (2004) 'Science and Technology - Tenth Report' U.K. House of Commons, Science and Technology Committee, July 2004, at

Hovav A. & Gray P. (2001) 'Managing Academic Electronic Publishing: Six Case Studies' Proc. 9th Euro. Conf. Infor. Syst. 27-29 June 2001, at

Howard M. (2001) 'Does Disintermediation Make Sense? Proc. 11th Nordic Conf. on Infor. and Documentation, Reykyavik, Iceland, 30 May - 1 June 2001, at

King D.W. (2004) 'Should Commercial Publishers Be Included in The Model for Open Access through Author Payment?' D-Lib Magazine 10, 6 (June 2004), at

King D.W. & Tenopir C. (1998) 'Economic Cost Models of Scientific Scholarly Journals' ICSU Press Workshop, Keble College Oxford UK, 31 Mar - 2 Apr 1998, at

McCabe M.J. & Snyder C.M. (2004) 'The Economics of Open-Access Journals' Preprint, 22 October 2004, at

McCabe M.J. & Snyder C.M. (2005) 'The best business model for scholarly journals: an economist's perspective' Nature 1 March 2005, at

Malone T.W., Yates J. & Benjamin R.I. (1987) 'Electronic Markets and Electronic Hierarchies' Commun. ACM 30, 6 (June 1987) 484-497

Morris S. (2005) 'The true costs of scholarly journal publishing' Learned Publishing 18, 2 (April 2005) 115-126

Nature (2004) 'Memorandum from Nature Publishing Group', Appendix 21 to HCSTC (2004), at

Odlyzko A.M. (1995) 'Tragic loss or good riddance? The impending demise of traditional scholarly journals' Intern. J. Human-Computer Studies (formerly Intern. J. Man-Machine Studies) 42 (1995) 71-122, and in the electronic J. Univ. Comp. Sci., pilot issue, 1994, at

Odlyzko A.M. (1997) 'The economics of electronic journals' First Monday 2, 8 (August 1997), at, accessed 26 March 2005

OJS (2006) 'OJS in an Hour', Open Journal System, version 2.1.1 of 30 July 2006, at

Page G., Campbell R. & Meadows A.J. (1997) 'Journal Publishing' Cambridge University Press, 1997

Rowland F. (1999) 'Electronic publishing: non-commercial alternatives' Learned Publishing' 12, 3 Page (July) 209-216, at

Rowland F. (2002) 'The peer-review process' Learned Publishing 15, 4 (October 2002) 247-258, at

Shapiro C. & Varian H.R. (1999) 'Information Rules: A Strategic Guide to the Network Economy' Harvard Business School Press, 1999

Smith W.T. (1999) 'The deconstructed journal - a new model for academic publishing' Learned Publishing 12, 2 (April 1999) 79-91, at

Springer (2005), at,11855,5-40359-12-115394-0,00.html

SQW (2004) 'Costs and business models in scientific research publishing' Wellcome Trust, London, 2004, quoted in Morris (2005)

Suber P. (2002) 'Open Access to the Scientific Journal Literature' Journal of Biology 1, 1 (June 2002), at

Suber P. (2003-) 'Timeline of the Open Access Movement' Earlham College, 2003-, at

Suber P. (2004a) 'Open Access Overview', at

Suber P. (2004b) 'Praising progress, preserving precision', SPARC Open Access Newsletter, issue #77, September 2, 2004, at

Suber P. (2004c) 'What is Open Access? An Overview' Proc. Conf. Open Access Publishing, Society for Scholarly Publishing, November 2004, slide-set at

Suber P. (2004d) 'A Very Brief Introduction to Open Access', 29 December 2004, at

Valauskas E.J. (1997) 'Waiting for Thomas Kuhn: First Monday and the Evolution of Electronic Journals' First Monday 2, 12 (December 1st, 1997), at

Walker T.J. (2002) 'The Future of Scientific Journals: Free Web Access?' LearnedPublishing 15, 4 (October 2002) 279-84, at

Willinsky J. (2003a) 'Scholarly Associations and the Economic Viability of Open Access Publishing' J. Digital Infor. 4, 2 (2003), at, accessed 26 March 2005

Willinsky J. (2003b) 'The Nine Flavours of Open Access Scholarly Publishing' J. Postgrad. Med. 49, 3 (July-September 2003), 263-267, at;year=2003;volume=49;issue=3;spage=263;epage=267;aulast=Willinsky;type=2

Willinsky J. (2005) 'Open Journal Systems: An example of Open Source Software for journal management and publishing' Library Hi-Tech 23, 4 (2005) 504-519, at

Willinsky J. (2006) 'The Access Principle: The Case for Open Access to Research and Scholarship' MIT Press, 2006

J. Willinsky & M. Ranjini (2007) 'Open access on a zero budget: a case study of Postcolonial Text' Information Research 12, 3 (April 2007), at


An early version of this paper was prepared in response to an invitation to address a session on 'The Impact of Open Access on Publishers, Librarians and Academics', at the Fiesole Collection Retreat Series, no. 7, Melbourne, 29 April 2005. An accompanying slide-set is available. The analysis and the presentations have subsequently been substantially further developed.

The paper was stimulated by Colin Steele, University Librarian of the Australian National University. It has benefited greatly from the assistance of Colin and of a number of journal editors and association officers, and from the comments of a number of colleagues and Fiesole community members. The evaluative comments, however, are mine alone.

Author Affiliations

Roger Clarke is Principal of Xamax Consultancy Pty Ltd, Canberra. He is also a Visiting Professor in the Baker & McKenzie Cyberspace Law & Policy Centre at the University of N.S.W., a Visiting Professor in the E-Commerce Programme at the University of Hong Kong, and a Visiting Professor in the Department of Computer Science at the Australian National University.

xamaxsmall.gif missing
The content and infrastructure for these community service pages are provided by Roger Clarke through his consultancy company, Xamax.

From the site's beginnings in August 1994 until February 2009, the infrastructure was provided by the Australian National University. During that time, the site accumulated close to 30 million hits. It passed 65 million in early 2021.

Sponsored by the Gallery, Bunhybee Grasslands, the extended Clarke Family, Knights of the Spatchcock and their drummer
Xamax Consultancy Pty Ltd
ACN: 002 360 456
78 Sidaway St, Chapman ACT 2611 AUSTRALIA
Tel: +61 2 6288 6916

Created: 29 March 2005 - Last Amended: 10 November 2007 by Roger Clarke - Site Last Verified: 15 February 2009
This document is at
Mail to Webmaster   -    © Xamax Consultancy Pty Ltd, 1995-2022   -    Privacy Policy