Pre-print (submitted for publication)

Pre-print (submitted for publication)

Selling the Right to Cite:

A Citation based funding scheme for high-profile, non-commercial publishers to operate profitable in a "semi" Open Access environment

Markus Schneider

Researcher, CTI/DTU

Markus@cti.dtu.dk

Abstract

How can a prestigious, academic journal offer immediate, free access rights to all its post-prints while at the same time being able to generate sufficient revenues to be profitable on a lasting basis?

The paper addresses this important question and proposes a citation based funding scheme for existing, non-commercial, prestigious journals. Such a funding scheme could - in combination with the currently applied peer review charging model - give this specific group of high quality journals not only an economic incentive to offer free access to their post-prints and to operate profitably on a lasting basis, but it would also lower the costs for the individual author/citer to reasonable levels and link the economic to the academic value of a post-print.

The paper outlines the proposal, shows its advantages over the current peer-review charge model and discusses the main objections against such a citation based funding scheme. The paper concludes that there are various advantages of the citation based funding model and argues that - especially due to the lack of economic thinking in academia and to the lack of institutional pressure to omit citations - a "pay-to-cite" business model is not as abstruse as it looks at first sight.

Introduction

At the outset, the author believes that there is a mutual dependency in the academic publishing market rather than an "academics-dependency" only. While publishers depend on academics as writers, peer reviewers, editors and readers, it is equally true that the latter heavily depend on (existing) publishers as well: namely for their personal career advancement. While money is not the main driver for academics in their daily work, prestige, honour titles, and research grants are. Generally, academics only can "buy" those "academic goods and services" (e.g. honor titles, research grants, prestigious jobs) if they publish. By publishing academics "earn academic money" (i.e. publications and citations) which they use to "buy" academic goods and services. Unlike self-archives and unknown, peer-reviewed, non-SCI (OA) journals, especially high-ranked SCI journals provide academics with the currency (publications and citations) they need to advance in their careers.[1] This "currency function" of publications is the basis for the academic reward system and the driving force in academia in general.

To illustrate this currency function of publishing: A young, unknown author writes an excellent paper and submits it to a suitable, but unknown OA journal (or PlOS for instance). The very same paper would have been accepted for publication in Nature. Hence, in order to maximise his investment (writing the pre-print in order to advance in his career) he needs to submit it to Nature because with the academic income he receives from the Nature publication he can impress any hiring committee. In other words, by submitting to a suitable, but unknown or low-impact OA journal the author is actually making a sacrifice (that this is hardly a good incentive and driver to promote OA is clear). The same holds true for mid-level and even senior researchers. There is fierce competition in the academic market and often there are many applicants for jobs, research grants etc. In those circumstances, any author is making a sacrifice by supporting unknown or low-impact OA journals if he could have published the paper in a top SCI journal. [2] Many academics submit their pre-prints first to the highest ranked journals (which would maximise their academic return on their time investment). If those journals reject, the author submits it to a lower ranked journal (and thereby lowering his return).

Currently, as only SCI publishers are in a position to generate the "academic currency", the author believes that there is a mutual dependency in the academic publishing market. In turn, this requires OA supporters and academics to accept concessions (unless the deeply rooted reliance on publications is replaced by alternative means of evaluation/matching supply and demand of academic goods and services) and to take the needs of publishers (operating profitably beyond the bare essentials) into account.[3] A citation based funding scheme would not limit the amount that universities need to pay to the costs of running a journal to the bare essentials; but the same amount of money would be spent more effectively and more to the benefit of the research community as a whole.

1 The Proposal: Selling the Right to Cite

Basically, rather than selling access rights or charge (to the individual) prohibitively high fees for peer-reviewing, journals can sell the rights to cite its papers:

Journal C makes all its publications OAI-compliant and immediately and freely available on its website. On any post-print a note is added: “All uses such as reading, copying, distributing, storing, printing and citing are permitted and free of charge. If this paper is subsequently cited by another publication obtaining a Digital Object Identifier (DOI), the sponsoring institution of the citing paper agrees to pay X Euro to C.”

Importantly, citations for all other not-publishing related purposes (working, research and student papers) would be free. Only if the citing paper is accepted for publication, the fee needs to be paid.

From a business perspective, Journal C offers a flat-rate scheme (for current subscribers and institutions with a lot of publication output and citations) and the possibility to "pay-per-cite" (for non-subscribers and occasional citers). To link revenues to costs and lower the price tag for a citation, journal C could also charge a peer-reviewing fee for papers which have been submitted to it.

In practice, this funding scheme is not going to be attractive to many journals, but only a few high profile journals which are cited often since they have a broad base to distribute the costs (citers). Moreover, commercial publishers which seek profit maximisation and need to sell their low impact journals are not going to implement such a funding scheme for their high-profile journals because of the very need to sell their low-impact journals.

While the non-commercial target group is small in terms of quantity, it is the top in many academic disciplines [4]. Hence, if successful the top non-commercial journals in any given discipline could offer immediate, free access to their post-prints and still operate profitably while the costs for "semi-OA" are allocated to those authors/institutions who benefit from other author's time and other institution's/publisher's monetary investment for their own career advancement.

2 Advantages of the citation based funding scheme

There are a number of advantages of the citation based funding scheme: unlike under the peer review charge model, the costs for running the journal are distributed among many authors/citers and therefore lower to the individual and easier to bear. In addition, the citation based funding model is more systematic and therefore very easy to implement. (rather than selling access rights, publishers sell citations rights). Moreover, from a more theoretical perspective the economic value of a paper is linked to the academic value.

(a) Distribution of Costs

A study conducted by the PNAS showed that 50% of authors are willing to pay a surcharge for immediate OA. Of those 50% around 80% are not willing to pay more than 500 for the peer-reviewing service. Less than 6% would be willing to pay more than 1500$ [5]. In contrast, the study says that 500 is a fourth of the amount that "might be needed to cover journal operations without subscription income". Also, it has been reported that non-commercial publishers would need to charge around 3000 per paper, [6] and - in the case of Nucleic Acids Research by OUP - 500 is heavily subsidized. [7] The Wellcome Trust estimates the costs for a high-quality journal to be around 1950 $ per paper. [8]

While detailed cost figures are not revealed by publishers, it seems obvious that there is a huge discrepancy between the willingness/possibility to pay (<500) and the required amount (>1500) making the peer-reviewing model unattractive to publishers. Hence, on the assumption that those figures somewhat represent the economic limits of authors and publishers and implicitly the likelihood of establishing a working business model for existing journals, the "peer review" fee model is not going to give - even non-commercial - publishers any economic incentive to make the switch to OA. It seems necessary to find an alternative way to generate additional revenues for the journal and to lower the costs for the individual.

Irrespective of any concrete number, at the one end in terms of cost distribution, any closed access journal distributes its total costs (and profits) among its subscribers. Generally, the higher the number of subscribers, the lower the costs for the individual recovering the journal's costs and profits. At the other end, the peer-reviewing model distributes the whole costs on a few writers whose papers are accepted. This results in the high figures (3000) that even non-commercial publishers need/want to charge in an OA environment.

A citation based funding scheme would lower the costs for the individual to bearable levels. A citation analysis conducted in an OA environment (ArXiv) showed that around 8% of all papers have been cited between 12 and 40 times, and 2% of all papers more than 39 times (Hitchcock, 2002). In other words, for those 10 %, the costs for peer-reviewing the paper can be distributed among 13+ users as opposed to just one author. While those figures should not be taken literally (as it will depend who is citing the papers - the same author/sponsoring institution; self-citations), logic tells that papers with a high number of citations have a broad base to distribute the costs.

(b) Existing Package Deals

From a practical point-of-view, another big advantage of the citation based model is that it is more systematic than the peer-reviewing model. Rather than negotiating with every university flat-rate, package deals [10] - which are necessary in order to reduce costs to bearable levels to the individual - universities which are already subscribing to journal C have already concluded such package deals. They would not pay for access, but rather for the right to cite for publication purposes (which implies free access to the post-prints). On the assumption that all universities which are currently subscribing to high profile journals also cite those journals, the flat-rate "citation charge" would be the same as now (The free rider problem can be addressed by introducing a reasonable fee for an individual citation).

A further advantage of citation based funding is the possibility to implement it fast. Publishers could make all their papers freely available within the next edition by adding a citation note similar to the one mentioned above. Importantly, since it is an existing journal, there are no additional costs for libraries and there is no transition period (this is a major problem with the strategy to establish new high profile journals to compete on pricing).

(d) Usage is controllable

Importantly, usage (citation) is controllable. Unlike other types of uses (e.g. copying, printing, storing at end user level), the formal requirement of obtaining a DOI and advances in citation analysis makes it easy to track citations and base a reliable remuneration system on it. The obtaining of a DOI can be conditioned upon the proof of subscription to J or payment of the citation fee. Citations for all other not-publishing related purposes (working, reasearch and student papers) would be free.

(e) Competition for "blockbusters"

If citation based remuneration is successful, it would not only give publishers an economic incentive to make their papers freely accessible, but perhaps it could even create competition for pre-prints of recognized authors. Lesk shows that readers read (and cite) pre-print papers on ArXiv primarily from authors they already know (Lesk, 1997). Hence, in combination with self-archiving of pre-prints on (filled) pre-print servers, publishers seeking to increase their revenues by publishing "blockbusters" could ask the author for permission to peer-review and publish the paper in the hope of peer-reviewing a "box office hit" generating many citations/income.

(f) Access to the most important research

While the absolute number of journals for which such a citation based model could work is small (in arXiv 10% of papers receive more than 12 citations), the top 10% account for the majority of all citations. Brody has found that in the arXiv 4 856 out of 291 033 papers (1.7%) account for 33% of all citations and 24 876 (8.5%) account for 66% of all citations.[12] Hence, while the number of papers that can be published with such a funding scheme is small, they represent the most important and valuable papers in their disciplines.

(g) Economic value is linked to the academic value

From a more theoretical perspective, citation based funding would link the economic (income; from the publisher's view) to the academic value (citation). Currently, there is a big pricing gap between top commercial and non-commercial journals in this respect.

3 Objections against the citations based funding scheme

Citations are extremely important in academia and fulfill various important functions (e.g. evaluation of researchers and departments, tenure and award of research grants and even in relations to salaries). Hence, it is necessary to consider and discuss any potentially negative effects. As it seems abstruse to charge for citing a paper, this section examines the most obvious objections in the context of the academic realities.

(a) Fewer Citations

The most obvious objection against a citation based funding scheme is that there will be fewer citations because the university needs to pay for it. For the following reasons, the author believes that a citation based funding model should not be rejected at the outset and deserves some more thought.

Would the author cite less?

First, the past decades and the serial crisis show that the individual author/reader does not think in economic terms when it comes to their academic writings. Academics who reject the idea of a "box office academic publication", academics who do not know/think/care about prices of the journals they use in their work and academics who first (and only) think of moral (copy) rights in relation to "their" creations (as opposed to the economic funds that are necessary to pay their salaries) and disregard rights of publishers and even (university) funding institutions paying for their research, illustrate that - on author level - there is often no economic thinking. In other words, it seems strange to assume that those ("typical" ?) academics outlined above are going to omit a citation because the university would need to pay for it.

In the opinion of the author, if the citation is necessary and improves his own work, the author is going to cite. Indeed, it is difficult to criticise or agree with another writer, to refer to studies backing up one's own point-of-view without citing properly. Also, one can assume that academics would not steal someone else's idea and commit plagiarism simply because the university has to pay for it. Generally, just like in any company, the employee does not care how much he is spending since he does not have to pay for it because it is work related - unless the company forces him to do so.

Would the university pressure to cite less?

So, would there come pressure from universities to save costs by omitting/avoiding citations? In the past, certainly pressure to save costs for serials came primarily from the funding institutions. But one should keep in mind that this pressure is primarily directed at commercial publishers who charge much more than non-commercial publishers and often offer less quality. (Bergstrom, 2001; Bergstrom and Bergstrom, 2004) Non-commercial publishers often charge very reasonable for their publications. The author believes that universities would not pressure to save costs when the value received is actually worth it (and is plausible to assume that the building on the high-quality work of someone else by citing him is "valuable" and "worth it" to justify the payment of citation fees).

Apart from the "cost-quality equilibrium", from a business perspective, if journal C offers a flat-rate to cite, universities which now spend X Euro for access to the journal would spend the very same amount for citing the papers. With a flat-rate offer, why should an university which regularly cites top journals cancel their subscription or pressure authors to use fewer citations? In this respect, for instance at NCSU, out of 718 subscribed RE journals the top 41 journals (which are not even the most cited SCI journals) received - on average over the last 20 years - at least 20 citations per year by NCSU authors.[14] If those 41 journals offer flat-rates (pricing is the same as now) why should NCSU cancel or pressure their authors to omit citations? Moreover, the university would not incur any additional costs as the access rights have been transformed into citations rights.

So, in the view of the author, in the case of the top non-commercial journals, the main question is whether universities without a current subscription are not citing because it would cost something. For journals wanting to provide free access to their post-prints, the important question is to what extent current non-subscribers are citing their journals in connection with the proposed citation based funding scheme. As citing is currently free of charge, authors who obtain journals through inter-library loans and cite them would indeed incur new costs. Three points should be considered in this respect:

First, as outlined above, the incentive of non-subscribers not to cite in order to save costs is limited by the non-economic thinking in academia and the need to cite in many cases (criticism, support, reference study, etc.). Second, the steady income stream of current subscribers enable journals to operate profitable and would allow them to be flexible in the case of developing countries for instance (this is based on the assumption that non-commercial publishers do not seek profit maximisation at all costs). Third, those journals even have a strong incentive to allow non-subscribers to cite: While being "generous" and allow non-subscribers to cite (freely or on favourable terms) does not generate (much) economic income, it does generate additional academic income (citations) which has an impact on the SCI ranking and the number of subscribers.

To summarize, in the light of the non-economic thinking of many academics, the lack of incentive to omit citations because of its costs-quality equilibirum, the assumed lack of institutional pressure to save costs at the expesense of high quality academic writings, and primarily due to the the possibility to offer flat rate pricing schemes, the author belives that there will not be fewer citations under a citation based funding scheme.

(b) Cancellations and Fee-Waiver Programs

A reasonable pricing scheme and an adjusted price tag for an individual citation could be a deterrent for universities to cancel their subscriptions. If the price tage for a citation is reasonable (high) it will not make universities switch to a "per-per-cite" scheme. As for non-subscribers (e.g from developing countries), the constant stream of income derived from current subscriptions could allow publishers to be generous and let them cite their papers on very favourable terms.

Another objection relates to citation based funding as another toll. But what is a "toll"? Currently, the author pays for the peer reviewing service. Under the citation model the author pays for peer reviewing its own paper and using other people's work. Fact is that peer reviewing and human quality control cost money and that the strategy to establish new OA journals is costly to libraries, very time consuming and difficult to implement in all academic disciplines.

In the view of the author, the important question is not whether there should be a "toll" - or even how high the total toll is - but rather how high is the toll to the individual and what are its effects on scientific progress. As for the latter, access toll actually restricts access; peer-reviewing theoretically limit the dissemination (of work by authors who do not have the necessary funds) and citation tolls theoretically limit citations for publications purposes. Whether those negative effects actually occur is another thing, impossible to predict without practical testing and will depend a lot on the business models.

(d) Pricing

Just as with the author publication model, the setting of a price tag (for a citation) is required. Total costs of the journal are - at least to the journal - known and journals can experiment with price tags for citations (and perhaps benefit from the impact of free access). If there are not fewer citations because of the need to pay for it (see above), the track history of citations allow publishers to estimate the price tag of a citation. In this respect, non-commercial publishers have shown in the past that they cite reasonably for the quality they offer.

(e) 30 Dollar limit

ScienceDirect is selling access (and citation) rights to their articles for 30$. Citation price tags for non-subscribers would most likely need to be higher. However, one cannot compare the two figures since the benefits from free access to all post-prints are much broader. For instance, not all articles that are accessed are cited for publication purposes. The costs for researching (accessing articles for other purposes than citing) are zero under the citation based model. In total, citation based funding is much more usage based than the access model.

(f) Bare Essentials

A citation based funding scheme does not reduce the costs for a journal to the bare essentials[15] and perhaps it would even allow some high-profile non-commercial publishers more profits than now. However, it should be kept in mind that it is commercial publishers who deserve the blame for the serial crisis and not the few top non-commercial publishers which charge reasonably for their quality. Even if non-commercial publishers would earn more after they make their post-prints freely available, this - in the opinion of the author - is acceptable due to the tremendous benefits of free access, the low costs to the individual, the high quality that those publisher deliver and - last but certainly not least - due to the current market realities with publishers as the essential link in the academic communciations process.

In addition, in the view of the author, it is irrealistic to assume that publishers will ever lower the costs to the bare essentials. Given the strong reliance on SCI publishers in particular, why should they? Even PlOS which was established to compete against high profile journals charges - initially - 1500$. It seems that this is the "bare essentials" and the price to be paid in todays academic publishing world and reflects the dependency on SCI-publishers. [16]

(g) Not True Open Access

From a terminological point of view, the citation based funding model is not OA as in most definitions because one needs to pay for a citation. In conformity with the definition and in order not to increase more harmful confusion about the term "Open Access", the paper does not claim that citation based funding is a business model for an OA journal.

However, citation based funding potentially could reach the main goal of the OA movement: (i) to eliminate all access barriers to academic information hindering the dissemination of scientific information and (ii) to give readers a broad range of rigths. Under the citation based funding model, there are no access barrier, and the usage barrier is never "visible" as citing for any purpose (other than formal publication) is free. Limiting the citation right for publication purposes - and gaining access, distribution, storing, citations for all other purposes, etc. - is in the view of the author worth it and - especially in the light of the strong/dominant position of publishers - a reasonable compromise.

Conclusion

Under the citation business model, the reader does not pay for access (and prevents others from having the very same access) nor does the uahot pay for peer-reviewing; under the proposed funding scheme, the author of accepted publications pays for citing other papers (perhaps in addition to peer reviewing its own pre-print). The model links payment to actual usage of the work. Citation based funding could allow high - profile non-commercial publishers to make all their post-prints immediately and freely available while being able to generate sufficient revenues for its operation without a transition period. Perhaps the biggest advantage of the model is that it is a systematic way to distribute the costs of running a high quality journal. In combination with already existing subscriptions, sound business models and flat-rate offerings in particular, the citation model could very well successfully adress the important issue of fewer citations.

The paper shows some advantages of the citation based funding model and discusses some of the most obvious objections. Regarding the latter, I believe that the outlined observations are sound and reasonable and deserve some more thought by others and are worth considering before rejection a "pay-for-cite" business model at the outset.

About the Author

Markus Schneider holds a law degree and is Researcher at the Danish Technical University. E-mail: markus@cti.dtu.dk.

Notes

[1] In the view of the author, one of the prime and most deeply-rooted problems of making OA successful is the reliance on the SCI. Unless there are found other ways to generate "academic income" used to buy academic goods, publishers are having a "central bank" monopoly for the production of currency and are therefore in a much stronger popsition as it appears at the first glance. See also Jean-Claude Guédon, 2001. "In Oldenburg's long shadow: librarians, research scientists, publishers, and the control of scientific publishing", In: Creating the digital future. Association of Research Libraries. Proceedings of the 138th Meeting, and at http://www.arl.org/arl/proceedings/138/guedon.html, accessed 30 April 2004; and Robert Parks, 2002. "The Faustian grip of academic publishing", Journal of Economic Methodology, 9(1), 317-335 and at http://econwpa.wustl.edu:80/eps/mic/papers/0202/0202005.pdf, accessed 30 April 2004.

[2] While publications and citations are not the only academic income, they are the most important source. Who can ever become an editor if he does not publish? Who is ever hired for research projects if he does not publish something on his past projects?

[3] The Failure of the PlOS letter which has been signed by more than 30 000 from 180 countries is evidence of the strong (even dominant) position of publishers. Also, the stock of ReedElsevier has been recovering substantially and recently even outperforming the FTSE-100 after various down-grades from investment banks in the second half of 2003. It looks as if the financial markets and investors are less worried about the future of ReedElsevier (and their profitable closed access business models) than in 2003.

[4] Ted Bergstrom, 2001 "Free Labor for Costly Journals?", Journal of Economic Perspectives, Volume 15, Number 3 pp. 183–198 and at http://www.econ.ucsb.edu/~tedb/Journals/mypapers.html, accessed 30 April 2004.

[5] Results of a PNAS author survey on an open access option for publication, Nicolas Cozzarelli; Kenneth Fulton; Diane Sullenberger, 2004, PNAS 3 February 2004, volume 101, number 5, p 1111 and at http://www.pnas.org/cgi/content/full/101/5/1111, accessed 30 April 2004.

[6] American Physiological Society estimates that it would have to charge authors $3,000 to publish papers in each of its 14 journals. The American Association for the Advancement of Science says it would have to charge $10,000 per paper in Science The Promise and Peril of 'Open Access', The Chronicle of Higher Education 30.01.2004, http://chronicle.com/free/v50/i21/21a01001.htm, accessed 30. April 2004.

[7] Richardson M.; Saxby C. OUP contribution to thew Nature debate on Open Access from 8.4.2004 at http://www.nature.com/nature/focus/accessdebate/12.html, accessed 30. April 2004.

[8] Costs and business models in scientific research publishing, Report by the Wellcome Trust, 2004 http://www.wellcome.ac.uk/en/images/costs_business_7955.pdf, accessed 30 April 2004.

[9] Steve Hitchcock et al, 2002 "Open Citation Linking: The Way Forward", D-Lib Magazine October 2002 volume 8, number 10 and at http://opcit.eprints.org/dl00/dl00.html/, accessed 30 April 2004.

[10] For instance, the University of Helsinki pays 6200 for having all their research published in any BioMedCentral journal.

[11] Lesk, M (1997) Practical Digital Libraries : Books, Bytes, and Bucks; Lesk, M; Morgan Kauffman

[12] E-mail conversation with Tim Brody.

[13] See for instance Bergstrom, 2001) T. Free Labor for Costly Journals?, Journal of Economic Perspectives, Fall 2001; Ted Bergstrom, Carl Bergstrom 2004 "The Costs and Benefits of Library Site Licenses to Academic Journals", Proceedings of the National Academy of Sciences, 20.01 2004) and at http://www.econ.ucsb.edu/~tedb/Journals/mypapers.html, accessed 30. April 2004.

[14] Those 718 journals are RE only journals. However, the numbers aim to illustrate that universities cite certain journals on a regular basis and it does not seem likely that there would be pressure to omit citations. (http://www.lib.ncsu.edu/colmgmt/elsevier/) (requires guest login).

[15] In the view of the author, rather than focussing on the total costs, the more relevant question is how to lower the cost for the individual in an OA environment (see above on the distribution of costs). A focus on lowering the costs for the individual - while accepting higher overall costs - has the important benefit that it creates an incentive for the publisher to support OA.

[16] In this respect and in connection with the current peer-reviewing fee model, Duke University published 4500 papers per year. Assuming a charge of 1500 per paper, the OA costs would be 6.75 million whereas total library spending (including humanities, medical centers and access to online databeses) was less (6.6 million). While package deal can lower the costs, it seems that the "introductionary offer" by PlOS of 1500 does not bring much of cost savings for research institutions with a lot of writing. Hence, one can raise the same objection (that citation based funding does not in any way reduce costs for research institutions) against the peer-review service fee. See The Promise and Peril of 'Open Access', The Chronicle of Higher Education 30.01.2004 (http://chronicle.com/free/v50/i21/21a01001.htm, acessed 30. April 2004.