- 1 What is a Business Model?
- 2 General remarks
- 2.1 Open Access (OA)
- 2.2 Toll Access (TA)
- 2.3 The Publisher vs. The Journal
- 2.4 Profit vs. Non-profit
- 2.5 The market
- 3 Implications of Business Models
What is a Business Model?
The concept of Business Models (BM) needs some clarification. It is a concept much spoken about, but more rarely defined. We will rely on Wikipedia for the description and definition following here. The article describes Business Model in this way:
[…]describes a business model as consisting of nine related business model building blocks. Thus, a business model describes a company's business:
- core capabilities: The capabilities and competencies necessary to execute a company's business model.
- partner network: The business alliances which complement other aspects of the business model.
- value configuration: The rationale which makes a business mutually beneficial for a business and its customers.
- value proposition: The products and services a business offers.
- target customer: The target audience for a business' products and services.
- distribution channel: The means by which a company delivers products and services to custom-ers. This includes the company's marketing and distribution strategy.
- customer relationship: The links a company establishes between itself and its different customer segments. The process of managing customer relationships is referred to as customer relationship management.
- cost structure: The monetary consequences of the means employed in the business model. […]
- revenue: The way a company makes money through a variety of revenue flows. A company's income.
These 9 business model building blocks constitute a business model design template which allows companies to describe their business model.
For a more detailed discussion of business models and how to work with them, see Osterwalder 2007.
Even if the concept of BM can be described in a few words, as above, applying this model on any given business will take amounts of work and intelligent analysis. A Business Model can in principle only be used on an actual business, like a scientific journal or a publishing house. Here we will focus on some implications of this model thinking on how Open Access journals must perceive their business – as opposed to that of their publisher, or TA journals – in general, trying to focus on factors that are common to OA journals as such, with emphasis on traits where OA journals are different from TA journals.
Open Access (OA)
There are various “flavours” of OA, and clarification is necessary in order to establish a foundation for this report. Open Access can mean anything from giving free access to read editorial content for the time being, to giving away all rights to exploit the content for perpetuity to the general public, for free.
Here, OA will be taken to mean Open Access as described in the Berlin Declaration:
Definition of an Open Access Contribution
Establishing open access as a worthwhile procedure ideally requires the active commitment of each and every individual producer of scientific knowledge and holder of cultural heritage. Open access contributions include original scientific research results, raw data and metadata, source materials, digital representations of pictorial and graphical materials and scholarly multimedia material.
Open access contributions must satisfy two conditions:
The author(s) and right holder(s) of such contributions grant(s) to all users a free, irrevocable, worldwide, right of access to, and a license to copy, use, distribute, transmit and display the work publicly and to make and distribute derivative works, in any digital medium for any responsible purpose, subject to proper attribution of authorship (community standards, will continue to provide the mechanism for enforcement of proper attribution and responsible use of the published work, as they do now), as well as the right to make small numbers of printed copies for their personal use.
A complete version of the work and all supplemental materials, including a copy of the permission as stated above, in an appropriate standard electronic format is deposited (and thus published) in at least one online repository using suitable technical standards (such as the Open Archive definitions) that is supported and maintained by an academic institution, scholarly society, government agency, or other well-established organization that seeks to enable open access, unrestricted distribution, inter operability, and long-term archiving.
This means that here OA journal will mean (unless the context makes it clear that it is meant otherwise or this is explicitly stated) a scientific journal published electronically on the internet with the full content being made available for free to anyone with internet access, giving away all rights to exploit the content further electronically, as long as the author’s right to be cited as author and being loyally quoted is not violated.
This is not to say that other definitions of OA may not be as valid as the Berlin definition, but the Berlin Declaration definition seems to be the most widely accepted definition in the OA community. Other definitions, whether explicit or implicit, often leave more business model options open to the publisher. While this may be of interest, in this analysis the most “extreme” definition of Open Access makes for the most interest-ing discussion.
Toll Access (TA)
Toll Access is the traditional model for scientific journal publishing, with readers having to pay for access to content. This is no longer synonymous with paper-based publishing, but as the general model seems to be to provide both paper-based and electronic access to content, TA journals are strongly modelled on the paper version and its affordances.
The Publisher vs. The Journal
This wiki will focus on the journal, not the publisher. One must bear in mind that the business model of the journal could be quite different from the business model of the publisher. One reason for the present discussion of OA might be just the fact that in many instances the business models, and business interests, of publishers, and those of scientific journals, might be in conflict with each other. The journal is, after all, the means for the publisher to reach his goals – while the publisher is the means for the journal to reach its goals. This is a kind of symbiosis that can be filled with conflict, at least when the goals of the journal and the publisher are at odds with each other.
Profit vs. Non-profit
One should also bear in mind that both OA and TA journals may be profit-maximizing businesses – or be interested in maximum distribution while breaking even, as to cover costs. There is no clear equation that says that OA equals non-profit (or altruism), and TA blatant profit maximizing, though one might see some tendency that OA journals seem less profit-oriented than TA. This might, however, be due to the fact that OA journals have not yet found the same constantly profit-creating business models as TA journals have.
The major OA publishers are businesses that will be just as profit-oriented as their TA counterparts, but they will have to find other means of generating that profit. And it might be possible that OA publishing models make the level of profit seen among some TA publishers impossible to reach. And it is also conceiv-able that the means of generating profits for an OA publisher will not be at odds with the goals of science and scientists to the degree that TA publishers’ business models seem to be.
When you analyze your business, you have to be clear about whether your goals are economic, i.e. that you want to make profits from publishing your journal(s), or whether your goals are dissemination within eco-nomic limits.
A major difference between TA and OA journals is that they operate in different markets, with markedly different economic traits. In the following, I will describe some of these differences, as an understanding of these differences is important in order to understand the economics of OA journals.
The market for TA journals
TA journals sell content to readers, or to representatives of readers. In a marketplace where there is free competition, sellers have to compete for buyers, and there is no room in an efficient competitive market for prices that yield high profits – such prices would make it interesting for more sellers to enter the market-place, driving prices down. TA journals do not compete in such a market. Where there are more than one journal that covers an area of science, readers cannot satisfy their needs by buying a journal at random, be-cause the content of one journal is no substitute for the content of another journal, unlike when choosing between different brands of e.g. soft drinks. The market for TA journals is termed a monopolistic market, because every journal is in a sense a monopoly in itself.
Another feature of the classical market is that there for any one buyer is a limit to how much one wants to buy, irrespective of price. This is a result of the (rapidly) decreasing marginal utility of consuming another unit of the good. If you want a bottle of soft drink to quench your thirst, your interest in yet another bottle will be marginal compared to your interest in the first bottle. As we all know, having access to one scientific journal does nothing to quench our thirst for access to yet other journals in the same field, to the extent journals cite one another having one will make one want the others. What limits the number of journals a library subscribes to is the library budget, and within that budget, journals are (mainly) chosen for their importance, not by price.
A seller in a competitive market has to accept a market price that he cannot do much to influence. In a monopolistic market, the seller has to consider his pricing strategies. If the seller lowers his price, he has to consider that the reduction may sell more copies, but at the same time, the price reduction will lower the revenue from the original sales volume. This has the effect that in a monopolistic market the market price will be higher and the volume lower, than in a competitive market.
The consumer surplus and price discrimination
In any market, competitive or monopolistic, where there is only one price for a product, there is a consumer surplus. Many buyers would have been willing to buy the product at a higher price. For each such buyer the difference between the market price and what the buyer would have been willing to pay, is a win for the buyer and a loss for the seller. The sum of these differences is the consumer surplus in the market.
If the seller could negotiate prices individually with buyers or groups of buyers, he could increase his profits by exploiting the consumer surplus. Such a differentiation of prices is called “price discrimination”. Price discrimination only works in markets where products cannot efficiently be traded between buyers; else, buyers who face low prices could buy more and sell their surplus to buyers who face higher prices.
A practical example of price discrimination is in the market for consumer electronics. New products are introduced at high prices. When consumers who are willing to buy at high prices have been exhausted, the prices are lowered a bit to reach new groups of consumers, this is repeated until the products are mainstream and sell at prices that to a greater extent reflect production costs. Such a model with prices varying over time cannot be used for subscriptions, though one could see selling back copies at a discount price as a variation of the model.
The effect of price discrimination is that sellers can sell more while still earning more than in a situation where you have a monopolistic market without price discrimination. With price discrimination, some buyers will face the same price as in a competitive market, a price that reflects the marginal cost of the seller for providing the good or service in question – again, a price that is lower than the seller could sell at without price discrimination.
In the traditional market for TA journals, price discrimination was limited to having different subscription prices for institutions, individual subscribers and student subscriptions. The cost of negotiating individual subscription rates would far exceed possible extra income generated.
The mechanism that allows TA journals to use price discrimination efficiently as a strategy to enhance their profits is bundling journals and selling bundles (or licenses to bundles) to institutions. While traditional pricing strategies would let all institutions meet the same subscription prices, making small and poor institutions pay the same as large and wealthy, bundling and negotiating allows publishers to tailor prices to institutions’ size and financial abilities. As letting buyers pay according to their financial ability will increase revenues without increasing costs, publishers’ profits are increasing.
The marginal cost of distributing a bundle of electronic journals is very small, thus can price discrimination strategies allow publishers to sell bundles to poorer institutions, e.g. institutions in poor countries, at virtually no cost to the buyer. Even giving away free subscriptions to bundles may be rational, because it both will create more citations for the publisher’s journals and create a demand that makes it possible to charge for subscriptions later.
The market for OA journals
One of the defining facts about OA journals is that they do not sell content to readers, they give it away. But there are buyers also in the OA market, otherwise OA journals would have no income. If we overlook institutional support as a source of income, the important buyers are the authors. They buy quality assurance and branding, author services and access to a readership through the journal.
While TA journals are small monopolies because the content of one journal is no substitute for the content of another journal, the quality assurance, branding, author services and readership of one OA journal are not substantially different from that of another journal. While OA journals, through various means, would like to brand themselves as being in some sense different from, and better than, other journals in the same subject area, these differences are less significant than the difference in content between journals in the market for TA journals. For an author, Journal A is an acceptable substitute for Journal B, even if it is not an exact simi-larity between them. Therefore, OA journals must compete in a market that more resembles the traditional competitive market. They have to compete for content through a combination of what they offer the author and what price they charge. Thus, OA journals do not have the same possibilities as TA journals for creating excessive profits for their owners.
The different markets for journals
One must conclude that there really are two different markets for journals. One for TA journals, which is characterized by monopolistic competition and excessive profits to publishers. Trading is to a large extent done through negotiations between representatives of readers (libraries/library consortia) and large publishers. The end-users, the readers, see little or nothing of the costs incurred.
The other market is for OA journals. This market is competitive, through differences in price versus service quality, and the costs are generally very visible and felt by the authors as end-users, who often have to pay the costs out of their research funds.
The two markets interact – the readers and the authors are the same persons, irrespective of market – and the total costs for journals are funded through the same institutional budgets. But there is a major problem: The real decision makers as to which model to use, which market to participate in, are the users in their capacity as authors. The authors face quite different costs in the two markets: None in the TA market, where costs are hidden in institutional budgets invisible to the authors, and article processing charges (often EUR 1250–2500) levied on the authors research budgets in the OA market. This “uneven playing field” where OA faces financial obstacles that TA do not face, makes for unfair competition between the two models of jour-nal publishing.
The efficiency caveat
One could easily be led to believe that the diminished possibilities for excessive profits in the OA model in itself would be an argument for its being more beneficial to science.
This only holds true if all other factors are equal between the two models. TA journals (at least the more profitable ones) are often published by large commercial publishers that have gained a lot of experience and competence in publishing journals. There are many reasons to believe that there are economies of scale in journal publishing, i.e. that the larger the volume published by a publisher, the lower the per article cost of publishing.
OA journals are generally published by smaller publishers, and all commercial OA publishers are relatively new firms – and OA publishing is different from TA publishing, meaning that reusing TA publishing experience in OA publishing firms does not necessarily provide maximum efficiency.
There is therefore reason to believe that costs in the OA publishing business is higher than in the TA publishing business, even if OA publishers face lower distribution costs than TA publishers do. If costs are higher, to a greater extent than profits are lower, OA publishing is not a profitable model for science at the present. Both size and competence will increase over time, given the present trends, decreasing relative cost inefficiencies in OA compared to TA.
Increased costs in the OA model could also be offset by the value of increased dissemination of scientific results through OA publishing – but this not make it financially better than the TA model.
There is reason to believe that over time, and as volumes expand and efficiency increases, OA will be economically better for science than today’s TA model – but we may very well not be there, yet.
The question of efficiency and economies of scale makes it imperative that OA publishing be organized not as individual one-man shows but in larger units, preferably as medium-sized publishing companies. Such companies should be large enough to be reasonably efficient, yet small enough not to be able to exploit market inefficiencies.
Implications of Business Models
With pressure from funding agencies, universities and, to a certain extent, researchers, publishing research articles under the concept of open access (i.e. so that they are available without cost to any reader) is a growing tendency. In fact, it would not be overly presumptuous to anticipate that in five to ten years most research articles will be published with open access. The uncertainty at present is how this will be accomplished. Technically, with the internet, it is easy to make an article available to anyone who wants to read it. However, there is much more to publishing a research article. Most significantly, it is still felt by most that an article should undergo an evaluation and editorial process before it is ready to be made available generally as a scientific publication. Publication of research articles also requires that these articles are easy to find and this usually means that similar articles are collected together and published as collections, still referred to as a journal and that these collections need to be indexed in databases. Furthermore, the articles must be stored in a safe way, with guaranteed accessibility over a long period of time. All of this means that it is not cost-free to make articles available for anyone to read them. For journals operating under the principles of open access then, they need to generate an income that allows their material to be available free of charge. Historically, journals have generated the income required to publish through subscription fees. In the early days of the open access movement, arguments were made that internet-based, open access journals would be much cheaper to produce and hence the subscription fees could be replaced with greatly reduced income streams from other sources. This however, is not necessarily true. Continuing to work within the paradigm of traditional academic publishing whereby articles are reviewed, edited and formatted prior to publication leads to the inevitable conclusion that the cost of producing an open access journal is not significantly less than for subscription-based journals. The question then is how to raise this rather significant income. A number of models have been suggested. At one end of the spectrum is what is often referred to as the author-pays model. Here the author(s) pay a fee for their article to be published (after it has been accepted for publication); this fee being set so that it allows all the expenses of operating a journal to be recovered. Actually to say that it is the author that pays is somewhat misleading since the money comes from the research grant that the research was conducted from or from the author’s institution. At the other end of the spectrum, business models based on sponsorship and/or advertising revenues have also been proposed. In between these two, lie models which rely upon multiple sources of income, the total of which are sufficient to cover operating expenses. One such example is what is called the hybrid model. Here, a journal continues to sell subscriptions to allow access to their material but gives authors the possibility to pay a once-off fee to make their article freely available to anyone. Similarly, some journals allow subscribers immediate access to material but make the same material freely available after an embargo period (of, e.g., 6 or 12 months).
The purpose of this section is to analyze the implications of these various models for operating open access journals for authors, publishers, libraries and research institutions. The report will begin with an analysis of the cost of operating a journal to give an idea of the necessary magnitude of fees for authors or income streams from sponsors or advertisers. With these figures it will be possible to comment on the impact they will have.
Running a Journal: the Costs
To begin with, it must be clearly stated that the purpose here is to give a flavour of operating costs and not to be a detailed financial analysis of journal production. Typically (or traditionally) the production of a journal requires taking a submitted article (perhaps even soliciting the article in the first place), organizing reviews, communicating with the reviewers and authors, making a publishing decision, formatting/editing the article, proofreading it and then publishing it (here “publishing” is used rather vaguely since it includes perhaps printing as well as making the article ready for the internet). Beyond the “producing an article” activities there are also support activities such as marketing, administration and other overheads, and distribution. Obviously some of the costs are variable, depending on the number of articles received, the number published, the length of the articles, and their complexity (e.g. the number of figures and references). Other costs are fixed, such as staff salaries, administration and overheads, and marketing. Waltham (2005), presents the relative sizes of these various costs, for a range of journal types. Notably, there is significant variation between different journals, however, to give an idea, typically fixed article production costs, fixed support costs, variable production costs and variable distribution costs account for 35, 30, 20 and 15%, respectively (this is for journals having both printed and electronic versions).
Quantitative figures on the cost of publication are rather difficult to obtain from commercial publishers, however, several studies in recent years (Greco and Wharton 2008; Waltham 2005; Velterop 2005; King and Tenopir 1998) provide some insight. This information together with experience publishing (smaller) open access journals at Linköping University are use to give some guidance as to the real cost of publishing. It must be kept in mind through the following that the variation from one journal to another in terms of costing is very large; it is not exaggerating to state that the difference in scaled costs (based on size of operation) can be as much as a factor of five between the most expensive journals and the cheapest. Middle-of-the-road figures on a per article basis (for electronic articles only), however, are 7500 SEK for the editorial process, 500 SEK for the publication, 500 SEK for distribution and 6000 SEK for overhead-type costs. This gives a total of 14500 SEK (but keep in mind that published values vary between extremes of 7000 and 30 000 SEK). These numbers were mainly from learned society publishers with journals producing 6 to 12 issues per year and 10 to 20 articles per issue. For university-based journals, a typical scenario has a journal producing 2 or 3 issues per year and perhaps 10 articles per issue. For these operations, there are generally two or three people directly involved in the editorial process, for a fraction of their total working time. Taking, as an example a journal with the equivalent of a senior academic and a junior person working 20% of their time each, then the costing (on a yearly basis) works out roughly as about 110000 SEK for editorial staff, 40000 SEK for proofreading (since English is the usual language of a journal and the majority of the editorial staff in Scandinavia have English as a second language, professional proofreading is usually required), 15000 SEK for pure publishing activities (the technical aspects of putting articles and issues on the internet) and 60000 SEK for supporting activities. This comes to some 12000 SEK per article, similar to the more commercial operators above. In the university-based example, however, one must keep in mind that a significant fraction of the time of the editorial staff is often not directly attributed to the running of the journal (i.e. departments or universities allow academics to “donate” their time to the production of the journal as part of their normal work duties). Creating a printed version of a journal obviously increases theses costs both on the printing and distribution sides. Typically one might expect an increase of some 5000 SEK per article.
As a rough guide, then, in an author-pays model, a journal would need to have an author pay around 15000 SEK to have an article published open access (assuming for now that all expenses are to be recovered from the authors). This, however, assumes that journals continue to operate exactly as the traditional subscription-based journals have. With authors paying to publish, the real cost of the publishing process becomes clearer to researchers and also gives them the chance to make more informed decisions about where to publish. The range of publishing costs noted above results for a number of reasons including the level of service provided by the journal (e.g. the amount of marketing/distribution they do, the amount of formatting, editing and correcting they do, the rejection rate…), the style of the articles, the extent of reviewing, etc. Under an author-pays scenario, authors can not only choose a journal based on traditional aspects such as timeliness of publication, perceived quality of the journal, access to appropriate readers, etc. but also on how much it costs (or more accurately, whether the value gain is in line with the cost). This of course opens possibilities for open access journals to operate a little differently from how traditional subscription-based journals have. An example is to increase the author’s responsibility in the process. Formatting and editing by journal staff can be reduced to a minimum, requiring this work to be done correctly by the author, prior to acceptance being given. A similar onus can be placed on authors for proofreading. There is no particular need for journals to do this type of work or to pay for it. If an article is judge to be poor with respect to language usage by reviewers then it can be sent back with a request for a professional service to be used. This may seem to place an unfair advantage on those using a second language for their publication, however, there are very many examples of people writing in a second language with excellent results: it becomes up to the author to choose whether they want to improve their skills or pay for someone else to take of this aspect of publication. The net result is that the author fee can be reduced somewhat and authors are left with the choice as to how much service they are willing to pay for.
The author fee can also be reduced through development of other sources of funding; sponsorship, advertising, membership in a society, for-profit sales of print versions of articles and issues of journals, etc. can also generate revenue which can offset the amount that must be recovered from the authors. For larger publishers, the cost of more expensive journals can be offset by higher-volume productions (Public Library of Science is currently trying this in an attempt to break even).
For hybrid models, the cost charged to the author to have their article immediately available is usually similar to the fee arrived at above under the author-pays scheme. A survey of a range of hybrid publishers (University of California, Berkley, 2008) shows a range from US$1000 to 3000 or about 8000 – 24000 SEK (this conversion is done at a higher exchange rate than at the current extremely low value). Hybrid business models also include delayed access and limited functionality. Delayed access just means that an article is only available to subscribers for a period of time (typically 6 or 12 months) after which it become available to everyone, without charge. Limited functionality means that readers are able to read an article without charge but may not be able to print it, they may only get access to an html version or a text-only version (no figures of linking of references), etc. In both the latter cases there is no charge to the author.
Implications of Author-pays and Hybrid models for Open Access
Given that the cost for the idea of open access publishing lies in the range 5000 to 30000 SEK per article and that this is to be paid by authors, the question arises as to what effect this has on authors, institutions, funding agencies and university libraries. To some degree the answer depends on whether the fee lies near the top or the bottom end of the range; 30000 SEK is a significant amount of money relative to many research grants. Since a significant amount of the drive towards open access is coming from funding agencies, in many cases this type of expense is legitimate to request in a grant application and so publication becomes an integral part of the cost of a project. It would appear, then, that author fees will not end up costing researchers anything. However, the total budgets available to many funding agencies have been shrinking in recent times and are not going to increase just because the cost of publication is going to become part of a project’s cost. Instead, this cost will have to off-set other project costs (or in essence, project budgets effectively become a little smaller). Furthermore, however, studies of authors’ attitudes have shown that a significant fraction of researchers are not willing to pay for publishing, a service that they have seen as being free historically. There are those that believe strongly in the ideals of open access but most look more practically at the issue and feel that they do not have the resources available to afford to pay around 10000 SEK for each article they publish; for a larger research group this can be an additional cost of in excess of 100000 SEK per year. In many academic environments the pressures of career development are strong. Researchers are expected to publish a number of articles per year and in recent times these articles must also be published in “high quality” journals and result in significant numbers of citations. The result is that for many researchers, particularly those starting their careers, where achieving tenure is crucial, ideals are placed at a lower level than survival; any research grant has to be made to go as far as possible and unnecessary expenses avoided. Paying author fees, either to a purely open access journal or to a subscription-based journal offering a hybrid scheme, is not going to be a high priority. Compounding this problem for open access journals is the impression that open access journals are of lower quality. In reality, there is little quality difference between open access journals and subscription-based journals. However, it often is true, at present, that open access journals have a lower status (but not in all cases): they are often newer, less established, smaller, are not indexed by, for example, ISI and as a result have lower (or no) impact factor. With this line of thinking, even reducing author fees for open access journals significantly is not likely to result in a large fraction of publications being published in open access journals. From an author’s point of view, the hybrid model is more attractive, since it allows publication in what are thought of as the main-stream journals with high impact factors and good indexing; effectively, one is still seen to be publishing in the “best” journals. What detracts slightly from this is the current discrepancy between hybrid model author fees and open access journal author fees; the former tend to be higher than the latter, but this is a choice that authors can make and at current most would choose the higher fee, if they had to make a choice.
Author fees, whether part of a hybrid scheme or for a fully open access journal are a hindrance for open access journals. Few researchers are really aware of the amount of money that is spent by university libraries to make their work available, at least to the academic community. It is important that this information been clearly taken to researchers, illustrating what each research article really costs. It is also important to point out that if a significant fraction of articles is published with open access then subscription fees charged to libraries should reduce somewhat and that this savings will be (or should be) passed on to them in some form or another. Possibilities can include programmes in which universities pay some or all of the author fees, reductions in overhead costs taken from research grants… These are discussed in more detail below.
Authors are also being presented with a third alternative and that is depositing their articles in an institutional repository. Here the cost to the author is usually zero, as it is a service run by a university or its library. Material in these repositories is generally freely available to anyone with an internet connection and hence, meets the criteria for making an article available open access. Most subscription-based journals allow authors to make the full text of their articles available in a non-profit institutional repository, often subject to the conditions that the authors’ final draft is used and perhaps not until an embargo period (6 or 12 months in most cases) has past. There is a growing belief amongst researchers that this alternative allows them to meet ideological considerations and university or funding agency requirements while not costing them anything except for a few minutes of their time. For the most, they are correct, however, the main caveat is that accessibility of material in repositories is not as broad as would be ideal. Anyone can gain access to the articles, however, they are often difficult to find in the first place. Articles in repositories are unlikely to be indexed with the usual big indices (Scopus, ISI, Scirus, etc.) because of a (justifiable) conflict with commercial interests and so a common method for finding research findings is not available. Repositories are likely indexed with search engines such as Google and Google Scholar and that undoubtedly provides wide access to material. However, as most who use Google know, the hit list often has 1000s of entries and few people browse past the first 20; an article must appear within this short list to be located. For institutional repositories to play a significant role in make scientific research available freely, this issue needs to be solved.
For universities as a whole, the concept of authors paying to publish their work or paying to make it freely available, is going to require a re-evaluation of some of the flow of resources within the university. Universities have a responsibility to society as a whole and part of that is to share the work and findings freely. Universities, hence, must (and generally do) support the open access concept. The question is how to support this activity. Historically, research was made available through journals and universities paid, through their libraries, to allow access to this. In most places, anyone could use a library’s facilities (as long as they physically visited the library, only university members were allowed full, distance, internet access to all the services). For a typical mid-sized Swedish university, the cost of this access is of the order 15 to 20 million Swedish crowns per year. For such a research intensive university (i.e. while having a large undergraduate component, still placing significant emphasis on research studies), it would produce, roughly, 1500 peer-reviewed research articles per year. Simplistically (ignoring any fees authors might pay such a page fees, reprint fees etc), the university contributes about 10000 SEK to the publishers for each article published. From a readers’ perspective, an average cost per downloaded article (dividing the subscription cost by the total number of articles downloaded) is usually in the range 20 to 25 SEK. If this reality shifts such that authors pay for the publication process, rather than the libraries then, one may simplistically think that the libraries subscription budget can be used to pay the authors’ publishing fees. To some degree this is true. The major publishers have all stated plans (more or less clearly and with greater or lesser degrees of complexity) whereby they will reduce their subscription fees in some relationship to the number of articles for which authors pay the open access fee. One certainty is that if all the authors of all articles pay the author-fee, subscription rates will not drop to zero, so the offset is not 100%. Secondly, with this type of scheme there is quite a long delay period over which universities are effectively paying twice for articles. As currently proposed, as an example, the number of articles published in 2008 for which the author pays the open access fee will be tallied in 2009 and then will be used in the calculation of the 2010 subscription rates. Eventually, an equilibrium will be reached in which the appropriate rate is charged since the rate of increase in the number of articles being published with the open access option will become zero. Until that point, however, universities will be paying an excess for the publishing process.
Within the university, a switch from libraries paying a limited number of subscriptions (with bundled packages, libraries need only negotiate with a relatively small number of publishers) to either the library or some other body in the university distributing funds directly to authors will incur increased administrative procedures and hence costs. A small number of universities (e.g. Univesity of North Carolina, University of Wisconsen, University of California (Berkley)…) are currently experimenting with libraries, usually together with the university’s office for research services, paying the author fees for articles which are published open access. The details vary a little, but basically, the university pays if there are no research funds available, usually up to a limit (examples include 8000 SEK and 25000 SEK, sometimes with a lower limit for articles published under a hybrid scheme). Authors are required to apply for this money and from this a decision is made as to whether to pay or not. It is unclear from the institutions trying this whether the decision is made purely on financial basis or whether there is some judgment as to the importance of the work, as well. In any case, it is fair to assume that the evaluation and decision-making process quickly becomes significant (perhaps even untenable) if one scales up to the scenario when virtually all articles become published open access. If one makes the assumption that journal subscriptions do decrease with increasing uptake of open access publishing, then it is going to follow that library budgets will fall to reflect this and that perhaps a more realistic method to pass this on to researchers is through a decrease in the overhead charges removed from their research grants. Of course a full economic analysis needs to be conducted, but one could imagine a reduction of around 0.5 to 1 percentage points would be realistic. For Swedish universities this could be a possibility, since overheads are removed from grants, even those from government sources. This is not true in all countries, where universities are not allowed to extract this type of money from public funding.
Another alternative for universities is to pay institutional memberships. As an example, a university can pay a yearly membership to BioMed Central, giving all authors a discount on author fees for journals published through them. For what they classify as a small institution (up to 1500 faculty and graduate students in chemistry, biology and medicine), a yearly fee of about 35000 SEK would entitle all authors of articles published through BioMed Central journals to a 15% reduction in author fees. If one takes the average fee as 10000 SEK (they range from some 17000 SEK down to 1000 SEK, depending on the journal, for those that charge), then a breakeven situation is reached if 24 articles are published in a year. Universities can also elect to pay higher membership fees and have all author fees waived. The membership fee is simply of a “pay on use” type whereby each year either in advance or after the fact universities pay the author fees for the articles published. The Public Library of Science (PLoS) has a similar scheme: universities can pay to become a member and then get a discount on author fees.
So what does it all mean?
In a holistic approach with current attitudes, it is doubtful whether authors publishing in open access only journals can really work. A microeconomic analysis shows that open access journals can be as successful as subscription journals, if authors are willing to pay to publish. A problem arises, however, in a system with a mixture of open access and subscription-based journals. In theory, if one assumes that the number of articles published does not increase dramatically just because there are more open access journals (in fact, it is quite likely that if authors have to pay for each article they publish, they may reduce the number that they publish and focus more on quality rather than quantity, or combine what might have been two shorter articles into one longer one), then if the number of articles published under an author-pays scheme increases, fewer go to the traditional subscription journals. While fewer articles available through the subscription journals should really lead to a reduction in the subscription cost, it is unclear whether there is really a driving force for this to happen. It is more likely that as the number of articles published with open access publishers increased, subscriptions costs would remain largely unchanged. The result is that for a university as a whole, the cost of publishing increases. This also makes it hard for universities to fund some kind of compensation scheme whereby authors’ fees are paid for them, since library budgets cannot be reduced significantly.
A hybrid scheme, whereby authors pay to make their article freely available through a subscription-based journal, leads to a clearer relationship between articles that have been paid for by authors and subscription rates. Since the setting of subscription rates is not a particularly transparent process, one cannot say that the subsequent reduction as allowance for paid articles will be crystal clear. However, the relationship is more direct than above and makes it easier for universities to argue and insist upon lower subscriptions. With reduced subscriptions, funding is available to help authors pay for their articles, perhaps leading to an increase in the uptake. At some point the question arises as to whether it is worth libraries continuing to pay subscriptions, if a significant fraction of material has been paid for by authors. Obviously at that point, libraries have a very strong bargaining position for ensuring that they get something useful for the subscription money that they spend. For authors, of course, publishing with the journals that they are used to publishing with, with established reputations is an advantage.
Under a hybrid scheme in which authors deposit their articles in an institutional repository once they are published in a subscription journal, subscription fees to journals will be unaffected. On the other hand, this alternative is cost-free to authors while still allowing their work to be freely available. The uncertainty here is whether articles will really be seen by a wide audience. With the current development of national and even continental portals to these repositories, this problem may be minimized.
Greco, A.N. & Wharton, R.M., 2008, Should Unviersity Presses Adopt an Open Access [Electronic Publishing] Business Model for all of their Scholarly Books?, ELPUB 2008 conference. King, D.W. & Tenopir, C., 1998, Economic Cost Models of Scientific Scholarly Journals, ICSU Press Workshop. University of California, Berkley, 2008, Selective List of Open Access and Paid Access Fees,  Velterop, J.M., 2005, Open Access Publishing and Scholarly Societies: A Guide, Open Society Institute, http://www.soros.org/openaccess/pdf/open_access_publishing_and_scholarly_societies.pdf Waltham, M., 2005, JISC: Learned Society Open Access Business Models, http://www.marywaltham.com/JISCReport1.pdf