Article processing fees or publication charges—charging contributing authors or (more typically) their proxies fees to subsidize processing and publishing costs—are among the most frequently discussed supply-side business model components for open-access journals. Article processing fees are based on the premise that authors and their host institutions are direct beneficiaries of publication in a scholarly journal. Article fees thus distribute a journal’s publication costs across those individuals and institutions that benefit most directly from a paper’s publication.

Almost half of open-access journals levy publication fees, which can include submission charges, page charges, illustration fees, and surcharges for color, and such fees represent about 30% of the revenue generated by open-access journals.

In the digital environment, where document length and color illustration have a minimal effect on costs, author charges tend to be flat-rate fees reflecting article processing costs. Some publishers levy charges on all articles submitted, while others apply them only to articles accepted for publication. As both rejected and accepted articles incur costs (some publishers report that rejected articles cost more to process than accepted ones), charging submission fees at the outset may reduce the number of inappropriate submissions that a journal must handle.

The level at which a publisher sets such fees will typically reflect a combination of its pre-press processing costs, its policies as to which submissions will incur a charge, the number of submissions, and the extent to which the publication charges offset actual expenses (in some instances, the publication charge may be intended to completely cover the cost of processing; in others, the charge may only partially defray costs). The latter will depend both on the publisher’s cost structure (and hence the level of the fee) and on the receptivity to such fees in the journal’s field.

According to one survey, article processing fees are wholly or partially subsidized, either by a research grant (34%), a foundation grant (5%), or by the author’s host department (8%) or institutional library (27%). The payment of such fees out of an author’s personal funds appears relatively low—about 5% across all open access journals. Most publishers using the model make allowances for special situations (for example, individuals without a host institution or from less developed countries), assessing lower fees or waiving fees altogether when no institutional subsidy exists. Society publishers often discount article publication fees for members, or waive them entirely. The effect of these policies on fee revenue must be taken into account when calculating article processing fees. Appendix B describes a simple method for computing article processing fees for either submitted or accepted articles.

Several research funding agencies have policies supporting the payment of article fees, and a growing number of academic institutions have established funds to cover all or part of the article processing fee for articles submitted by affiliated authors to open-access publications. ,

2.1.1 Suitability

For journals in disciplines for which they have been a long-standing practice, article processing charges provide a logical model to support open-access dissemination. However, authors in disciplines without an established page-charge tradition (the case in much of the humanities and social sciences) may be expected to resist the practice.

As noted above, article processing fees seek to recover publication costs from the individuals and institutions that benefit most directly from a paper’s publication. While this is one of the advantages of such charges, it is also one of the principal objections to them when they are paid by the individual author seeking professional advancement. If a publisher attempts to transplant the model to a discipline without a tradition of page charges, positioning the charges as being paid by academic institutions, funding agencies, and other sponsors might prove less objectionable. Relying on such author proxies might allay author objections and dispel the perception of vanity publishing. This would also require encouraging sponsors of social scientific and humanistic research to implement policies to fund open-access publication that are similar to those in place for many scientific research funders.

2.1.2 Discretionary Open Access

Another variation on article processing charges makes payment of an article processing fee—and the availability of the article via Open Access—subject to the author’s discretion. Besides opening access to some of a journal’s content, this approach might provide a transition strategy, potentially even for disciplines in which article fees are not customary. Under this model, articles by non-participating authors may remain available only to subscribers or may be made available free online after an embargo period.

Some publishers realize substantial revenue from reprint sales to corporate buyers. Rather than assuming that such revenues would be lost under an open-access model, publishers that currently enjoy these revenues may want to reposition reprint fees as corporate sponsorships to support Open Access. Corporate purchasers may be willing to continue paying a fee where the publisher provides the reprint in a format that associates the sponsor with the publication’s identity and reputation.

2.1.3 Transitioning from a Subscription Model to Article Processing Fees

Subscription-based journals that wish to transition to open-access distribution funded by article processing fees may find themselves at an impasse: while they require a compensating income stream to offset the print journal cancellations that might result from open-access availability of the journal, some of their contributing authors may not have access to the funds necessary to pay for publication charges, and levying an article processing fee puts the journal at a competitive disadvantage. The severity of this problem may vary by discipline, depending on the policies of funders in a given field and by the practices of comparable journals.

An individual journal acting unilaterally to implement article processing fees might put itself at a competitive disadvantage in attracting author submissions. To mitigate this risk, a group of comparable journals—for example, society journals in a particular field or subfield, might transition to the new model in concert. The extent of this collective action might be limited to coordinating the timing of the transition, or it could include coordinated planning and marketing of the new approach. In addition to coordinated action with similar journals, the discretionary publication charge approach described below can help subscription-supported academic journals overcome these issues and manage financial risk in order to transition to open-access distribution funded largely or entirely by article processing fees.

In a transition to a discretionary model, a journal would lower its subscription price over time as the proportion of authors electing to pay the publication fee increases. In this way, a journal would be able to gain new revenue to support a transition to Open Access at a rate commensurate with the acceptance of publication charges by the journal’s author community. The transition period afforded by the model would also provide the time and context necessary to:

• Demonstrate the benefits to authors—including greater visibility, accelerated citation, and higher impact—of Open Access. At the same time, authors who cannot (or choose not) to pay publication charges can continue to publish in the journal.

• Allow researchers to begin to build publication charges into their research funding proposals, and for granting agencies to implement publication charge policies to fund publication of the research that they support.

• Assess the effect of the model on other journal income streams. Some of these, such as offprint and permissions income, might be expected to decrease. Whether other income streams, such as advertising, increase or decrease will depend on the manner in which they are implemented in an online environment.

Appendix C provides a financial forecasting template for introducing discretionary article processing fees and illustrates how discretionary fees might be implemented over time. The financial forecasting model takes into account the issues that need to be considered when planning such a transition; namely:

• Journals with relatively high publication costs will typically require higher publication charges. The point at which the article publication fee becomes a barrier to author submissions will depend on both the discipline (that is, the extent to which authors are accustomed to such fees) and the journal’s reputation (that is, the value authors perceive in publishing in a particular journal).

Of course, other income sources may allow a journal to set its publication charges at a lower level. Conversely, where possible, setting the publication charge slightly higher than the actual per-article cost can provide an operating surplus to offset years in which the journal incurs an income shortfall due to the timing of subscription cancellations and/or lower than anticipated author participation.

• Most publishers will find it difficult to project with confidence the rate at which authors will opt to pay publication charges. This suggests that a market study of the journal’s author community, supported by an effective promotional program, will help improve author participation, increase predictability, and lower risk.

• Where author participation rates are low, slow to grow, or fluctuate considerably from one year to the next, a journal may face short- and mid-term operating losses. For example, were author participation in the discretionary article fee program to drop from the previous year, the lowered journal subscription price might not generate sufficient income to cover the journal’s expenses. In many instances, it would be difficult to increase the subscription price sufficiently from year-to-year to cover sharp fluctuations. To reflect this, a journal might impose a cap (for example, 10%) on year-to-year price fluctuations.

• There may be a tipping point for cancellations of paid subscriptions that precedes the journal’s ability to support itself through publication charges. In come cases this could be addressed by setting the publication fee higher from the outset of the transition. As long as the initial publication fee is not so high as to discourage author participation, this could generate a surplus reserve that could be applied to subsidize the journal through years when it incurs an operating loss. Again, a journal should assess its risk by analyzing its author community thoroughly before undertaking the transition.

While such a transition strategy is not without risk, the risk can be anticipated and mitigated, especially by established journals in disciplines with a strong tradition of publication or page charges.

2.1.4 Article Processing Fee Examples

Prominent examples of publishers relying entirely on article processing fees include:

• BioMed Central (http://www.biomedcentral.com)

• Hindawi Publishing Corporation (http://www.hindawi.com), located in Egypt, leverages a low cost base to provide economically viable, labor-intensive publishing services. Hindawi publishes over 150 open-access titles.

• Journal of Medical Internet Research charges a non-refundable submission fee for regular articles. It also allows authors to pay an additional fee to accelerate the publication process. ()

• MedKnow Publications, of Mumbai, India, publishes over 80 open-access journals. (http://www.medknow.com)

• Molecular Diversity Preservation International (MDPI) journals, while headquartered in Basel, Switzerland, operates in China (http://www.mdpi.com). MDPI journals include Molecules (launched in 1996), the International Journal of Molecular Sciences (launched in 2000), and Sensors (launched in 2001).

• Optics Express, published by the Optical Society of America, sets article publication fees based on the length of the article (http://www.opticsinfobase.org/oe/submit/review/pub_charge.cfm#opex)

• Public Library of Science (http://www.plos.org)

The SHERPA/RoMEO site (http://www.sherpa.ac.uk/romeo/PaidOA.html) provides a list of publishers that offer optional paid Open Access. Selected examples are listed below. Peter Suber and Caroline Sutton have also analyzed society open-access journals, including those using article processing fees.

Discretionary Open Access Examples

• Examples of author-discretionary models from society publishers include:

• The American Physical Society, “Free to Read” option (http://publish.aps.org/FREETOREAD_FAQ.html).

• American Society of Limnology and Oceanography, the “Free Access Publication” program allows authors to make an article freely available immediately upon payment of a fee that supplements the journal’s page and color charges. (http://www.aslo.org/lo/information/freeaccess.html)

• Mineralogical Society of America, charges on a per-page basis (http://www.minsocam.org/MSA/ammin/e-pub_policy.htm).

• Royal Society of Chemistry, “Open Science” (http://www.rsc.org/Publishing/Journals/OpenScience/FAQ.asp)

• Royal Society, “EXiS (Excellence in Science) Open Choice” (http://royalsocietypublishing.org/site/authors/EXiS.xhtml)

From university presses:

• Cambridge University Press, “Cambridge Open Option” (http://journals.cambridge.org/action/stream?pageId=4088&level=2#4092)

• Oxford University Press, “Oxford Open” (http://www.oxfordjournals.org/oxfordopen/)

From commercial publishers:

• BMJ Publishing Group, “BMJ Journals Unlocked” (http://adc.bmj.com/info/unlocked.dtl)

• Nature Publishing Group (http://www.nature.com/press_releases/greengold.html)

• Springer Open Choice (http://www.springer.com/open+choice?SGWID=0-40359-0-0-0)

• iOpenAccess, Taylor & Francis (http://www.informaworld.com/smpp/authors_journals_iopenaccess_about~db=all)

• World Scientific Publishing, “WorldSciNet Open Access” (http://www.worldscinet.com/authors/openaccess.shtml)

Several publishers have instituted models that package article processing fees in an institutional subscription: in some instances, a fixed fee covers all manuscript submissions from researchers at the subscribing institution; in other cases, the institutional membership fee is based on the number of article submissions.

• BioMed Central offers both volume-driven and fixed-discount programs with pre-payment and payment in arrears options, with additional discounts for pre-payment (http://www.biomedcentral.com/info/instmembership.asp)

• Hindawi Publishing Corporation uses a flat rate (http://www.hindawi.com/memberships.html)

• Journal of Medical Internet Research (http://www.jmir.org/cms/view/Support_%2526amp%253B_Membership)

• Public Library of Science (http://www.plos.org/support/instmembership.html)

• The ScientificWorldJournal (http://www.thescientificworld.com/TSW/main/Static.asp?menuid=246&jid=141#)

Some publishers allow an institution’s subscription fees to cover discretionary open-access article publication charges for articles from authors affiliated with the institution. Examples of this approach include:

• National Academy of Sciences, discounts publication fees for articles from authors at subscribing institutions (http://www.pnas.org/site/subscriptions/open-access.shtml).

• Oxford University Press discounts optional article processing fees for authors from institutions that subscribe to the journal. (http://www.oxfordjournals.org/oxfordopen/charges.html)

• Springer waives the article fee for optional Open Access for some institutions and consortia, including the Dutch library consortium (Universiteitsbibliotheken en de Koninklijke Bibliotheek); the University of Göttingen; and the University of California system.

Some society journals waive article processing fees for members of the society; for example:

• Plant Physiology, published by the American Society of Plant Biology, does not charge publication fees to ASPB members (http://www.plantphysiol.org/misc/ifora.shtml)

When commenting on this page, please detail your experience with the model in question. The comment area is moderated and reserved for evidence- or experience-based discussion and requests for support in experimenting with different approaches.