SUBSCRIPTION SUBSIDIZED OPEN ACCESS
& THE CRISIS IN SCHOLARLY PUBLISHING
Note: critique of the green rights movement that appears below was elaborated at a conference Nov. 12, 2005 at Lehigh University. See presentation with notes here. There are also pdf versions of the slides and accompanying notes at the conference website.
Coverage:
(i.) Alternative publishing models to resolve the crisis in scholarly publishing. A number of models are discussed below; here is a summary list:
NOTE:
Brian Simboli brs4@lehigh.edu; (610) 758-5003.
CONTENTS
1. What is this webpage all about?
What is the crisis in scholarly publishing? "Journal prices are not dropping, and academic library budgets are not rising" (from http://www.libraryjournal.com/article/CA516819 ). Unjust pricing schemes -- unjust in the sense of the tradition of just pricing that emanated from the medieval period, if not prior -- are common in scholarly publishing by commercial entities, even if not universal. As a result, currently scholarly publishing is in a state of stasis, a game theoretic situation in which affected players wait for others to take the next move to resolve this crisis.
Quite possibly the whole enterprise of commercial scholarly publishing will implode on itself. This might be salutary. However, let us hope that less disruptive, more orderly transitions to a more viable system will soon be forthcoming. French Revolutions are ugly affairs.
When I broached the ideas on this page to a person who works in information technology and is intimately familiar with the Internet, he said, in effect, that scholarly publishing needs to catch up with the way things are done on the Internet generally.
The Internet generates a demand and expectation that material will be freely accessible; meeting this expectation is compatible with providing enhanced services of various kinds. Along these lines, the "subscription open access model" developed on this webpage is one way worth exploring to get out of the current quagmire.
Who might develop this model? While the only examples of the model known to the author of this webpage are from a commercial sources, it seems that universities are especially suited as implementers given, e.g., the cost-recovery basis on which it may well have to be implemented. Scholarly societies that do not use their publishing operations as means to fund non-publishing activities would also be suitable implementers.
(i.) "Subscription Subsidized Open Access"
There is much talk about open access these days. (See here for some resources). Open access discussions have failed to explore how "for-pay", subscription access can be made fully compatible with provision of open access. That is, there seems little recognition in many of the current debates that "open access"and "subscriptions" need not be diametrically opposed. To the extent that "open access" has turned into a sort of ideology that focuses unduly on issues of literature accessibility rather than affordability, it has helped to divert attention from this crucial datum.
This webpage explores how the concepts "subscription" and "open access", which to many will sound self-contradictory (or hopelessly oxymoronic), may well be a viable approach to overcoming the current crisis in publishing. The focus on the webpage is especially subscription subsidized open access as it pertains to ejournals, but the overall model pertains to other document types as well, such as online encyclopedias, or online books, or whatever.
Understanding the broad contours of the model is key before thinking hard about very concrete implementations. Below, I also suggest possible concrete implementations of the broad ideas discussed on this webpage, but only as a spur to discussion about ways to realize practically the more theoretical/conceptual aspects of the ideas involved.
How does subscription subsidized open access address the crisis in affordability of library resources? Precisely by making resources open access, it guarantees that margins charged by publishers for "premium access" to resources will be slim.
In subscription subsidized open access : all material is freely accessible; making it freely accessible keeps costs for premium access low; premium access charges make possible the very existence of the resources--including payment for editorial processes, peer review administration, and quality control--that is, the very things that make publishing a costly operation. That's the essence of the model.
One variation worth mentioning is that the model is compatible with author payments for publication, where those payments would supplement revenue streams from subscription payments. A variety of other variants of the model are discussed later on the webpage.
(ii.) Subscription Subsidized Open Access: Electronic Journals
It is well-known that journal prices have spiraled upwardly in price, far outpacing the ability of libraries to maintain their subscriptions. This has had significant opportunity costs in terms of ability to purchase monographs as well as print and online resources. The author of this webpage must routinely disregard advertising for many valuable resources because they are not affordable in the current situation. It is striking how the high cost of serials is cannibalizing the commercial publisher market for non-journal library resources.
While I am interested in the concept of subscription subsidized open access generally, the particular focus (though not exclusively) for now of this webpage is the "subscription overlay journal", designed to address the problem of affordability of ejournals. This is one species of the more generic concept, "overlay journal", that is receiving increased attention. To this author, the only known examples, thus far, of a subscription overlay are journals made available by BioMed Central (see below). (Incidentally, this was unknown to the author at the initial inception of this webpage.)
The model of overlay journals that is the focus of this webpage (the subscription variety) probably differs from the one that most followers of debates about publishing immediately assume when they hear the term "overlay".
In the thicket of considerations and counter-considerations that follows below (and is necessary to put flesh on the idea), it is easy to lose sight of the basic concept, which is simply this.
A subscription overlay journal is one for which users access a publisher's table of contents on an institutional or individual subscription basis. The articles so accessed are otherwise made available -- by the publisher -- for free access at either at some repository or repositories or on its own server. This model is distinct from approaches that call for authors to "self-archive" articles otherwise published by an ejournal publisher; here, the publisher plays no direct causal role in making papers freely available, other than to permit or even encourage authors, of their own volition, to post articles at repositories.
For people interested in variants of this approach, see the section about variants of the subscription overlay toward the end of this web site. But do not tread there until reading the other material. There is also the question whether subscription overlays is always the best term; perhaps "subscription subsidized open access ejournals" is a better term for certain uses. Again, see the material toward the end of the webpage.
This webpage is designed to summarize points made in listserv contributions and to spur further discussion and critique. No assumption is made about the viability of this approach, but the case is made below that it has various strengths. Whether those strengths are defeasible remains to be seen.
2. What is subscription subsidized open access?
Subscription subsidized open access electronic resources are those where the content is made accessible for a cost at the publisher's website, but where the material is otherwise made available by the publisher for free viewing. The electronic resources can be online journals, encyclopedias, books, or whatever.
This differs from other another approach, in the realm of journals, that calls for author self-archiving of articles, under which the author is permitted to archive an article at a repository.
Subscription open access is designed to provide a business model to make possible the existence of open access electronic resources.
References:
https://mx2.arl.org/Lists/SPARC-OAForum/Message/1652.html
In relation to subscription subsidized open access as it relates specifically to journals, one motivation for it is that, in many cases, current commercial pricing of ejournals is excessive and therefore unjust. (Here, the tradition of reflection on the "just price", with its roots in the medieval period if not before, is relevant.) To what extent pricing of other types of electronic resources is unjust is an open question. A second motivation is that there are certain goods that should not be subjected to an undue level of "commodification". Intellectual production falls in this category. This is a long story.
One assumption that does not animate this model is anti-business, anti-market sentiment. However, given the two motivations identified above, and the overall failure of the commercial realm to offer low-cost options, there is a very strong argument for non-profit organizations to develop business models that compete with the high pricing imposed by commercial entities. Whether commercial entities can succeed in this model is an open question. The model proposed here does not presuppose either a cost-recovery or a profit-motivated model. Whether it is solely compatible with a cost-recovery model or whether it can also reliably generate profits as well remains to be seen.
The "subscription overlay journal" is just one way of implementing subscription subsidized open access in the realm of journals. For discussion of non-overlay subscription open access ejournals, see the question below, "Is the term "overlay" really of any significance in the "subscription journal overlay" model? Why not just speak of subscription subsidized open access ejournals?"
For now, the focus will be on the subscription overlay journal idea, for two reasons. First, universities and colleges should take the lead in experimenting with this model, since many society publishers and very few commercial publishers will be willing to adopt this model. Second, the development of institutional repositories by universities and colleges lends itself to implementing this model.
Now, on to the specifics of the "subscription overlay journal".
The "overlay" is a type of journal that has been receiving increased attention. For present purposes, the overlay journal will be defined as one that provides a table of contents from which users can access journal articles as deposited in a publicly accessible institutional repository (IR). This definition is silent on particular issues such as these: whether one must pay for a table of contents to see the material, which is freely available on the IR; whether the material is or is not peer-reviewed; whether the finalized copy of a paper was or was not loaded in preprint form on that IR; whether or not the IR loading of articles is on a centralized IR or a distributed one.
Given that definition, overlays divide into two main types. Overlay journal is the "genus"; the species are these:
a . Pure OA journal overlays: overlay journals both of whose table of contents and articles are freely accessible. (Support for this model can derive from a variety of funding sources.) There has been a fair amount of discussion of this type of overlay and it may well be what is typically assumed in discussions about overlays.
b. Subscription journal overlays: overlay journals in which an institution or individual pays for the right to view a publisher's table of contents for a journal (and, if available, the full aggregation of tables of contents for a group of journals), and where each article is made otherwise accessible for free --by that publisher -- (concurrently with their being made available on the pay-basis) on a freely accessible institutional repository or repositories, or on the publisher's own server. Access to articles therefore can be via the table of contents or via use of search engines such as OAIster or Google.
The focus of this website is mainly "b". The economic advantages of this model in helping to overcome the crisis of library resource affordability will be clarified below. For now, two notes.
First, further differentiae are possible of the "subscription overlay" concept, but not to complicate matters, they will be listed at the end of this webpage.
Note that (b.) is compatible with authors paying some nominal amount as a supplementary revenue stream to subscription revenues. See below.
Second, in what follows, a particular type of "subscription overlay" is presupposed, namely, one that involves traditional processes of peer review. Reasons are given later on this webpage.
References:1. A description of the two varieties of the overlay concept, including a description specifically of the subscription overlay, is provided in my materials posted here. (Note: those materials refer to toll access overlays; a better rubric is "subscription overlays", since toll access may connote payment by individuals on an article-by-article basis, whereas the model described below would involve access to a table of contents either via institutional or individual subscriptions.)2. https://mx2.arl.org/Lists/SPARC-OAForum/Message/1565.html
3. For examples of overlay journals associated with arxiv.org, see here.
Well, here is just one possible scenario among many other possible ones. The focus here is universities and colleges or their consortia. (While anyone can adopt this model, there are reasons to think it is best suited for these entities. See elsewhere on this webpage.)
Heads of libraries, probably large libraries within a large consortium, or even libraries within a consortium of consortia, decide to pool their resources to create a clearinghouse office that maintains an "aggregator" portal to a large number of ejournals. A no-frills warehousing sort of operation would be the order of the day. Armies of student employees could be enlisted to support in implementing these projects and computer science graduate students given research assistantships to assist in implementing and maintaining the technical infrastructure. The goals would be initially modest: test the viability of this approach for a small stock of journals, which could be journals of society participants, or new journals, or journals whose editorial boards have elected to participate and which "defect" from their current publishing platform.
Initial founder donations or other sources of money would be secured to jump-start this model. (See question devoted to this issue).
This could but need not be done in relation to university presses. (May there be an argument for proceeding independently of them if they are not interested in working operations on a cost-recovery basis, if the latter is what is required to make subscription open access work?)
Articles would be maintained at an institutional repository such as arxiv.org, or at a centralized repository maintained by the consortium. All the articles would be open access, freely accessible for view by the public at that repository.
The university aggregator portal web site would provide all sorts of enhanced access to the journal content. (See the list of such services elsewhere on this webpage.) That is, institutions or individuals would pay for the right to use all the services at the webpage, including access to articles through journal tables of contents at the website. However, again, the public need not buy a subscription to view the content. All content would be freely accessible, and otherwise accessible on an enhanced basis via the consortium-maintained website at a subscription cost.
The model could be one like Project Euclid or BioOne or commercial aggregator sites, which serve as portals that aggregate access to a wide variety of journals, except that all the journal articles would in this case be freely accessible to the public on a non-enhanced access basis.
Who would the "publisher" be in such a scenario? There are two possibilities here.
(1.) First, the content might derive from journals whose editorial operations subsist independently of the university-led aggregation operation. Those journals would make their materials available on an enhanced access basis only through the university aggregator portal, i.e., the latter would be the sole, exclusive provider enhanced access to the articles. "Enhanced access" includes access to the table of contents plus the other enhanced services identified elsewhere on this webpage.
The table of contents could be assembled independently by the journal's staff or, better yet, would be assembled by the university aggregator, so that it has the same format as other tables of contents for the other journals made available by the university aggregator. If the independent journal staff assembles the table of contents, it could maintain that table of contents and have the university aggregator link to it, or provide the table of contents to the university aggregator for its own mounting and maintenance. If the university aggregator assembles and maintains the tables of contents, it would first receive the actual articles from the independent editorial operation. However this is done, in either case the table of contents (and associated enhanced services) would only be accessible via the university aggregator's portal website on a subscription basis.
The university aggregator would likewise, immediately upon making articles available in this way, make them freely accessible for 'non-enhanced access' at an institutional repository. If the content derives from journals whose journals editorial operations subsist independent of the university-led aggregation operation, then "publisher" focally and primarily describes those independent journals, and only secondarily or in an extended sense would the "publisher" be the university-led operation.
The challenge here would be for the university aggregator to find enough journals willing to participate in this way. Those journals whose editors view them as revenue-generators may well be deterred from participating in this model.
(2.) However, if the university aggregator aggregates some number of journals that it itself publishes, then it is in a primary sense a "publisher". Here it would be a matter of creating new journals of sufficiently high quality at the get-go that they will attract further quality research. The peer review services would be coordinated by the university aggregator, which would also put up the table of contents for the journal (for subscription access) plus make other enhanced services available, plus deposit the articles immediately on publication in a freely accessible, non-enhanced service repository.
Subscription revenues would be cycled back to the clearinghouse office operation (and if there are independent editorial operations involved, to those operations as well), which would be operated on a cost-recovery basis, at least initially.
Note, however, that the model on this webpage is compatible with either a cost-recovery model, or a profit-motivated, one. At the outset, universities would want to keep goals modest--namely, cost-recovery. The goal for them would primarily be to counter excessively high prices imposed by commercial entities, not to turn profits.
Revenues would fund: the clearinghouse office operation, plus payment for logistics of the editing processes incurred by a journal. The revenues could also be used to fund the IT costs for implementing, maintaining, and developing the tools and infrastucture that enable the enhanced features (identified elsewhere on this webpage), though university "founder" donations or grant funding would likely be needed to implement fully the initial incarnations of these tools. Otherwise there would be a sort of chicken and egg problem. (In this context, see David Stern's message here.)
Are universities in a position to handle the accounting and business processes involved in implementing this publishing model? The business and accounting issues cannot be any more complicated, and probably are a lot less complicated, than those associated with all sorts of business commitments and involvements that universities already have. Many universities already have their own presses, for example. Whether those presses as they are currently run are the best launching for new initiatives along the lines of this webpage is an open question. Nonetheless, the point remains that universities have a history of involvement in publishing efforts.
In the message linked to above, David Stern posted a question about the funding source for the repository. In my view, it is best to launch this model on the basis of a large, already existing repository such as arxiv (which is mentioned in the message). Revenues generated from subscriptions would best be funnelled into peer review system, not the repository infrastructure, at the outset. Only after there was a guaranteed source of income from subscriptions (possibly combined with author charges) would it be worthwhile to address how that revenue stream might be funnelled back to help pay for the repository.
The participating journals would maintain their traditional model of peer review, at least at the outset. (See discussion of this elsewhere on this webpage.)
What if the money generated by subscriptions did not suffice to fund these operations? In this case, supplementary monies would be needed. These could derive from nominal author charges (lower than the ones currently charged by open access author payment journals); funding organizations; or society funds garnered from membership dues. Again, initial seed monies would likely be needed.
There are obviously many practical difficulties and risks associated with implementing this model. However, it is hard to envision any other open access business model, other than the charging of high author charges, that would have a chance of challenging the hegemony of commercial publishers. The important point is to explore new business models, and even mix and match various approaches, to see what works. At the outset, this can be done with a small set of journals surrounding a specific subject, that can serve as a pilot.
6. What might jump-start the development of this approach?
If universities implement "subscription subsdized open access" (whether with respect to journals or other information types), one way to fund the initial business expenses involved would be to request seed money "sponsor donations" from universities and colleges.
Such sponsors could be offered reduced rates down the line on subscription costs and reduced author charges, if the latter are incorporated into the model.
Other sources: government funds merely on a seed basis?
[see other material at this website, including section 2 about subscription subsidized open access and section 4 about subscription overlay journals]
(1.) Something that at first glance --but only first glance--looks like the subscription overlay may already being practiced in the realm of physics publishing. If the information here is correct, many physics peer-reviewed postprints are being archived on arxiv.org and are therefore freely accessible. This is somewhat like a subscription overlay in its effect, since items are available both from the publisher and as archived at arxiv.
However, it is important to notice that the journals involved here are not subscription overlays. *It is critical to the subscription overlay model that the publisher ensures that all the articles it publishes are also available in its own open access repository or the open access repository or repositories elsewhere.* Therefore, the afore-mentioned facts about physics publishing do not constitute a de facto subscription overlay. (In fact, it apparently is just the green approach, criticized elsewhere on this webpage.) This is not mere quibbling about definitions. For unless the relevant physics societies themselves make the postprints available to arxiv for immediate publishing, they are relying on authors to self-archive their articles there. The drawback is that this does not guarantee 100% open access to these articles, since it is up to authors whether they decide to archive materials.
(2.) The following journals published by the commercial entity BioMed Central (BMC) are subscription journals that publish research that is open access, immediately: Critical Care, Genome Biology, Arthritis Research and Therapy, Breast Cancer Research. The four journals are: (a.) ones for which all the research articles are made available for free from each journal's own website plus the BMC website generally, plus the research articles are deposited immediately at PubMed Central and (b). ones for which reviews, feature articles, and other content of the journals are available at PubMed Central after a two year embargo period. They are not cases in which there is subscription payment for enhanced access if this is defined as including but also going beyond subscription-paid access to a table of contents, and where the current material is also freely and immediately available via other means (e.g., availability via PubMed Central). (See discussion below of variant models.)
References:
8. What are the economics of subscription journal overlays as opposed to other types of OA initiatives, including pure OA journal overlays?
Subscription subsidized open access for all document types (journals, books, reference materials, or whatever) is designed to address the crisis of library resource funding and to promote open access. In making items open access, it guarantees that margins charged by publishers will be slim.
(i). The subscription overlay model, which is the general model of subscription subsidized open access as applied to ejournals, would depend for its success on a publisher's ability to provide the following services effectively via the website for the journal(s); these are examples, and not necessarily an exhaustive list of services; note that paid access to a table of contents is not a necessary feature of the model:
- institutional "branding" that identifies library and other support services for the subscribing institution
- well-designed interface enabling searching of keywords, subject terms, and abstracts, across various journals
- easy integration with widely used course management softwares
- capabilities to integrate with federated search engine systems
- "if you like this, then you will like this", or "related articles" suggestion feature
- interface that can be individually tailored to include just the resources of interest to a researcher (not just email alerts, but also ability for a researcher to manipulate the webpage so that only resources of interest appear)
- sponsored links that appear when using search interface to locate articles
- journal article rating service (as in BMC Faculty of 1000)
- 'lo, even capabilities for downloading text to palm pilots.
(ii) In addition to revenues for standard sorts of enhancements, might revenue be sought in technical support services for implementing and using these value added features?
(iii) Might revenues be sought in advertising? This is a possibility. Note: it is interesting in this context that BioMed Central provides advertising via Google; see the top of its webpage, with its reference to "Ads by Gooooooogle" [sic]. One suspects that the amount of revenue that can be generated via advertising at a a publisher's overlay journal site (whether of the pure OA variety or the subscription variety) would be marginal, as it does not have the breadth of audience of (say) a Yahoo. But this should be studied, if it has not been already, especially with respect to overlay journals generally.
(iv) Might revenue be sought in sponsored links that show up when one does a search, as in Google?
(v.) Business synergies could arise between providers of such enhanced services. For an extra amount, one could get access not only to the services of one provider, but expanded and integrated coverage with other providers as well. So for example there could be sharing of cited/citing reference data between providers of these enhanced services and increased pricing for accessing that shared data.
(vi) The subscription overlay involves a business model that squares with the way libraries have done things for a very long time. That is, it squares with the way journal operations have been funded traditionally, namely, via subscriptions, the resulting revenues of which bankroll the publishing operation.By contrast, other OA models rely for their business support on: author payment for publication; money from benefactors; print subscriptions; special library funding; or government sponsorship. (Here, the reference is not to self-archiving initiatives, which rely on the largesse of publishers in allowing authors to self-archive articles otherwise available on a pay-basis through the publisher. In my view, this "green" approach is problematic, for reasons adduced elsewhere. )
Of course, these other OA business models can be mixed and matched with the subscription open access model. E.g., author payment (at a low rate, say) could serve as an additional source of income to supplement monies from subscriptions. An institutional subscription could bring with it a reduced author fee.
Here are strengths of the subscription overlay:
(a.) Unlike these other models, its funding would not be contingent on the vagaries and vicissitudes of other revenue sources. By contrast, built into the subscription overlay model is a method for cost-recovery, namely, payment by institutions for subscription-based access to a table of contents for one journal, or a set of journals.
(b.) The model can be executed on a cost-recovery basis, and can obviate (in one incarnation) commercial involvement that seeks acquisitive profit margins. It is does not, however, necessarily obviate all commercial involvement. When material is available for free on an article by article basis, institutions will be reluctant to pay some huge amount to see a table of contents or aggregation service. There is a threshold of pricing for the subscription overlay, below which institutions will be willing to pay to see an organized table of contents (or access an aggregation service), and above which they will not. That threshold will be much lower than the current pricing of the big commercials. This is for the simple reason that people will not pay for really expensive "enhancements" if they can get to the material on a repository for free. (See section about thressholds.)(c.) In those cases where OA journals require the author to pay for publication, there is a strong potential that this variety of OA will replicate the currently spiraling upward costs of the old subscription model. That is, the commercials will charge the author (or his/her institution) more and more. Who foots the bill? By contrast, the subscription overlay is designed to be low-cost and, in fact, if it is not so, it will not survive.
(d.) There has been a great deal of interest in institutional repositories. Their existence provides a basis for developing the subscription overlay model. While the author self-archiving model described as "green" is in my judgment not viable (see below) as a way to combat the problem of affordability--nor is it designed to be--it perhaps has helped spur development of institutional repositories by virtue of giving one more rationale for developing same.
(e.) Even if ultimately the pure OA variety of overlay journal fully establishes itself, the subscription overlay can play a transitional role in helping to establish a significant niche for overlay journals in general (regardless of their species). This is important because the funding models for pure OA overlay journals may be in flux for some time.
References:
9. What possible relationship could the subscription overlay model bear to author-payment OA models?
A mitigated version of the subscription overlay model would be one in which revenues are generated via enhanced services, and where authors contribute a small amount to make up for whatever shortfall in revenues there might be. This small amount may turn out to be lower than the current author charges imposed on author-payment OA models. Institutional subscriptions could provide decreased author charges as an added stimulus for selling and maintaining subscriptions.
There is nothing to suggest that mixing and matching various models cannot be attempted.
In the context of discussions of alternative publishing models, there is much talk about thressholds or equilibrium points that define critical transitions.
In the case of this model, an important transition point would be reached if one of the following were to tranpire; this assumes that monies for initial development would be forthcoming from founder sponsorships, grants, etc.:
Which of these three definitions of success proves definitive cannot of course be answered a priori, independently of rolling up sleaves and testing out various forms of the model.
One important symbolic or thresshold or milestone indicator of the success of alternative publishing generally and not just the model of subscription subsidized open access, is: when increases in subscription costs annually imposed on libraries equal rises in the CPI (consumer price index).
11. In summary, what incentives would an institution have for participating in this model?
1.Ease of ability in using the enhancements provided by the implementer of the model and perception that the value they add is worth the cost and that the implementer is keeping up with trends in IT.
2. For those implementers that impose them, institutional subscription licensing that affords lowered author fees (for authors at that institution) could also provide an incentive for institutions to stay on board.
3. As implementers increased, synergies could emerge between them. Just for example, provision of services in a coordinated way, such as batching (by various implementers) of their combined citing/cited data. Access to this data would incentivize subscribing institutions to participate.
4. What role might the good of institutional cooperation, in itself, as an incentive for institutions to continue participation in this model? If institutions perceived that subscriptions involved paying a bit more than what they perceived to be the value of accessing literature on an enhanced basis, might they be willing to foot that differential because they perceive the value of institutional cooperation? That is, might there be a price they would be willing to pay to perpetuate a cooperative publishing model in which many of their peers are participating--especially from the standpoint that this would be better than reverting back to a commercially dominated market for journals? This is an open question.
5. Heather Morrison suggests a possible incentive for geographically clustered universities or colleges to participate in this model in this listserv posting. Comments in response here.
Here are some key criticisms (first see email links immediately above).
(i) Why would any institution pay for access to content that is otherwise free?
Payment for content that is otherwise freely available is not unknown to librarians. Here is an example. The success of a such a product depends on whether librarians will find that it adds sufficient "value-added" aspects that it is worth it to them to pay for the service.In the case of subscription overlays, the idea is that institutions may be willing to pay for the enhanced features identified above. In a hybrid version of the model, revenues from subscriptions would be supplemented with other sources of revenue, such as nominal author charges.
(ii) Wouldn't an alternative service emerge that would reconstruct and provide access to the material otherwise available at the website?
The possibility that the table of contents can be reconstructed and therefore accessed in some way other than through the publisher's own website does not seem a defeating consideration, for these reasons:
- A possibility here is to build into the author's copyright agreement a clause on acceptable use stipulating that the article may be made available on a for-pay basis via only those enhanced services sanctioned by the journal publisher
- Perhaps portions of the metadata (exclusive of the some minimal metadata, such as the author's name) could be withheld for a period of time, a kind of embargo period, and then inserted after that embargo period. (A non-searchable publisher's "watermark" could appear in the pdf or html versions of the document whose metadata was thus embargoed, to indicate their provenance.) However, the reasons above suggest why it may not be necessary to engage in such a metadata embargo. If it were necessary to engage in such a metadata embargo, it is not clear that this would be problematic in a way that makes the model undesirable. The trade-off of embargoed metadata for a viable and sustainable business model, one that makes possible the very existence of a peer-reviewed journal, may well be acceptable.
An objection to witholding metadata, in accord with the last point, is that this goes against the whole grain of open access and OAI initiatives. Perhaps, but consider too that we are now in an environment in which many electronic journal publishers impose a several month embargo on access to newly published materials. What is more desirable: this state of affairs, or an environment in which subscription overlays are available, which makes the content immediately accessible, even if it is not as easy to get to as it would be if one has a subscription? Again, however, it is not clear that the viability of the subscription overlay model depends crucially on witholding metadata for an embargo period.
The definition given above of subscription overlays does not commit to a peer-review process of article evaluation. However, there is a very strong argument for subscription overlays to do so--at the outset--or for that matter any other model that wants to challenge the hegemony of commercial publishing.In a nutshell, to repeat a point already made: French Revolutions are ugly.
If reform of publishing is to occur, it is my view that it will need -- initially -- to emphasize lines of continuity with traditional ways of doing things. Whatever its drawbacks, traditional peer-review at least has the strength that it has largely been accepted as the standard route to ultimately publishing material that is "finalized". It is likely best, in establishing the infrastructure for alternative publishing models that will challenge current high-pricing, not to become unduly "creative" in testing new approaches that challenge the old way of doing peer review.
That is, for now, the best way to go about implementing overlay journals, and subscription overlays specifically, is no different from the way peer review is currently done. That is, an editor would assign peer reviewers, who would then evaluate articles following the guidelines that are currently used in publishing.
New approaches can be tested later, once the crisis of affordability is resolved. In this context, build on past practices, then challenge them. The sentiments here are along Burkean lines. An orderly transition to a better system is better than a chaotic one.
BMC makes available some subscription overlays. This is a for-profit organization. However, there is no reason why subscription overlays cannot also be produced by universities or societies or libraries.
It seems that the model is best suited for universities or colleges or their consortia. First, they have the most to gain from a fiscal standpoint, and would likely be better suited than other entities to promote and sustain a cost-recovery approach, or at least an approach in which the margins gleaned would be relatively slim, by comparison with the needs of other types of entities. Second, universities and colleges are implementing the infrastructural basis-"institutional repositories"-that provide (just) one possible incarnation of the model. As used here, these are repositories of publications or data or dissertations hosted by universities or other associations, where the material contained in the repository is freely accessible ("open access"). One instance of the model that these repositories make possible is overlay journals.
A few comments about possible promoters:
(a.) As for university publishing operations, they would have to approach this with the understanding that this is a cost-recovery model, not a profit-making one. Whether current university publishers would be willing to engage such a model is of course unknown. They may well not be willing to assume the requisite risk.
(b.) As for societies, those societies that regard publishing activities as a way of funding other non-publishing society activities will likely not be interested in this model. They will likely not be willing to assume the required risk-taking nor will they want to jeapordize their funding of society activities. Philosophical questions arise here about whether societies should help the crisis in pricing by adopting, even more so than in the past, low-cost publishing initiatives, even on a merely cost-recovery basis.
(c.) As for libraries:
(i.) We already one example of a library that has become involved in publishing. Consider, too, that Project Euclid was started by Cornell University libraries. Perhaps libraries, which already in the current environment have become quite savvy technologically, could pull this off. There is no reason to think they cannot. Consider the technical capabilities of just a mid-sized university in accomplishing digital projects such as these. Whether they can handle the business and logistical dimensions would remain to be seen. However, consider whatever university libraries make services available to corporations or the public for a fee; these would provide examples of how libraries can become involved in business operations.
(ii.) After some test runs on a small scale, large aggregation would enable an economy of scale that might enable universities to pull off a significant, long run commitment to subscription open access. It is perhaps *only* with such large aggregation that sufficient revenues could be generated to sustain this model. Also valuable would be consortial participation of university libraries, again to achieve these economies of scale. Participation of a large "IR", such as arxiv.org at Cornell, could be of value here. Various implementers of enhanced services like those identified elsewhere on this webpage could cooperate in providing cooperative pricing models in which subscribers pay more for utilizing not just one implementer's services, but for integrated access to the services of other implementers as well (just for example, cited/citing capabilities).
(iii.) Universities and their libraries can benefit from software that is available for journals publication management; this would take care of one dimension of the logistics. Example of such a software.
15. For what subject areas of publishing would the "subscription journal overlay" be most appropriate?
It's hard to think of why it would not be appropriate across the board. But this is an open question, meriting study.
16. What relationship does this model bear to the "green" approach involving self-archiving efforts by authors?
A subscription overlay journal is one for which users access a publisher's table of contents on an institutional or individual subscription basis. The articles so accessed are otherwise made available -- by the publisher -- for free access at either at some repository or repositories or on its own server.
This model is distinct from the green approach. In the green approach, the publisher plays no direct causal role in making papers freely available, other than to permit or even encourage authors, of their own volition, to post articles at repositories. Thus to the extent that articles are freely available in the green approach, this is a contingent matter dependent on the good graces of publishers and the willingness of an author to self-archive.
By constrast, the subscription journal overlay model -- by definition, and in se-- makes articles freely accessible that are otherwise accessible via for-pay subscription access to a table of contents.
The so-called green approach has perhaps played a role in spurring the development of IR's and contributed to awareness about alternative publishing models. However, these side-effect benefits are merely a circuitous route to the real issue at hand: making, as a focus, development of new business models that more directly address the crisis in periodicals pricing, rather than engage in the bibliographically confusing and redundant "double-publishing" that the green approach promotes. (As such, in my view the ability to self-archive is not a good justification for developing institutional repositories, which can serve other purposes, including providing of infrastructure for a reform of scholarly publishing.)
The green approach is marred, in my judgment, by a number of difficulties. The viability of the subscription overlay approach does not ride on the following critique; however, this critique suggests that more effort be placed by universities on developing overlays, rather than on pursuing the green approach, if it comes down to an either-or economic decision about what course to pursue, realizing that not all universities or institutions do not have unlimited staff support.
Concerning the green approach:
(i.) It sees too much distance between the issues of affordability and access to literature, when in fact these are tightly related issues. Quite simply, high costs make it difficult to expand access not only to journal literature, but to other types of literature such as online or print reference works and monographs. (See the opening comments of this webpage.)
(ii.) It makes accessibility to journal literature dependent on external factors such as: largesse of publishers, willingness of authors to self-archive; and a desire for institutions and governments to mandate that research results be archived by their producers. Also, aside from the issue whether many institutions will in fact mandate that their researchers archive materials (otherwise made available by publishers in duplicative fashion), the question arises whether even if an institution does create this mandate, researchers will follow suit and self-archive materials consistently.
Institutional mandates do not seem as problematic as government ones (see iv.), but there are daunting logistical hurdles, administratively, in assuring that busy researchers, or disorganized ones, put all their material in archives. Who is going to "police" them--department secretaries or librarians? Will the latter have to run comprehensive searches of online dbases to make sure that everything published is also being self-archived? Also, can we can count on widespread development and maintenance of training programs to educate people on location about self-archiving? Will someone have to sport a baton or truncheon to get researchers in line who do not care that much about self-archiving? Also, note that no research institution is going to fire a highly productive researcher who does not care about self-archiving, so such institutional mandates would not have teeth. (Perhaps the culture of institutions could change in the future so that this becomes a requirement of achieving tenure, but what about all the research output of post-docs or non-tenured researchers working at universities or researchers working at non-university research centers.)
(iii.) It does not promise the existence of a stable infrastructure for the future, given that it depends in "after-glow" fashion on the existence of a stable of journals for which institutions are currently paying high costs. Even if its stated goal of achieving 100% open access to ejournal articles was achieved, the system on which it is based would remain unstable and dependent on publisher extension of green rights, and/or government support for such a system, and/or continued participation of researchers.
(iv.) An independent, philosophically tinged point: with respect to requirements by governments that research results be archived, this may involve a surfeit of government intervention, in violation of what is known in social philosophy as the "principle of subsidiarity". (No assumption is made here whether government funding of research is also in violation of that principle; the latter in certain cases may be necessary in ways that mandating of self-archiving would not be.)
(v.) The concept of "subscription subsidized open access", including its application to ejournals, gets at the root of my criticism of the "ideology" of open access, which has diverted attention from the way in which for-pay access can be made fully compatible with provision of open access. This, in my judgment, is one of the negative consequences of the self-archiving ("green") movement, despite its possibly otherwise salutary effects in promoting more widespread access and in (perhaps) having played a role thus far in spurring development of institutional repositories.
18. What are further differentiae (sub-species) of subscription-subsidized open access for journals?
Variant models and hybrids of various kinds are conceivable. No commentary is offered here on their viability. In addition to below, consider that there can be various feedback loops between implementations of the model for various reference types. E.g., an encyclopedia or online book publishing on this model could be used to fund journal operations on this model, or vice versa. Or the model as implemented for one type of information resource (journal, or book, or encyclopedia, e.g.) could be used to fund implementation of another type of information resource that relies on a different model.
To reiterate from above, for convenience of reference here is the definition of the subscription journal overlay:
b. Subscription journal overlays: overlay journals in which an institution or individual pays for the right to view a publisher's table of contents for a journal (and, if available, the full aggregation of tables of contents for a group of journals), and where each article is made otherwise accessible for free --by that publisher -- on a freely accessible institutional repository or repositories, or on the publisher's own server. Access to articles therefore can be via the table of contents or via use of search engines such as OAIster or Google.
Imaginable sub-species are these:
(1.) In each of these two cases, the table of contents would (by virtue of the definition of the subscription overlay) be accessible at the publisher website only for pay:
(2.) Another distinction:
(3.) So far the assumption has been that the table of contents is one of the enhanced features for which subscription payments would be made. Here, a fee is charged for access to special enhanced services, where those services include access to the table of contents. Another model would involve a journal in which one can access the table of contents, which would link to open access articles directly off the journal webpage, but where enhanced services of other kinds would be "for pay". Making the table of contents available on a for-pay basis however might be necessary to sustain the revenue stream. It would depend on user perceptions about how much the various types of enhancements add value to a journal or journal aggregator website.
(4.) There is another variant model, wherein one pays for some content that is otherwise not available immediately, but where other content is also available for free, and where the subscription toll to the former helps fund the latter. This is a case where one pays to see material immediately that is not otherwise available unless one pays the subscription. The user pays for content not otherwise freely available, except say after a delay. The thrust here is more on the traditional subscription model where one pays for accessing content not otherwise available, and not on payment for enhanced access to material that is freely available. (I'll not speculate whether and how this model relates to the BMC journals mentioned above, even if those journals did inspire thinking about the variant model just described.)
(5.) Another model would involve a combination of (4) *and* payment for enhanced access to the content that is available for free.
19. Is the term "overlay" really of any significance in the "subscription journal overlay" model? Why not just speak of subscription subsidized open access ejournals?
Overlay in the context of ejournals refers to "overlaying" a table of contents "over" a repository of articles. On the other hand, if a publisher were to make articles freely accessible on its own server, would we want to call that server a "repository"? Such a use departs from the way "repository" is being used in current debates.
Here is a way to capture the nuances, terminologically.
Subscription subsidized open access ejournals divide into:
(a) subscription subsidized open access journals whose articles reside on the publisher's server, and
(b.) subscription overlay journals.
"Subscription overlay ejournals" then is a term used in the specific context of the model as implemented on institutional repositories and not a publisher's own server.
Advantages of centralizing the content to one large repository (for a given subject), and therefore for retaining the overlay aspect, are these:
(1.) There may be value in having the articles posted on a widely recognized repository rather than a publisher's own server. The former would provide one more set of search methods for getting to the content, plus there might be certain economies that come with many participants in the model posting everything centrally on this model.
(2.) From the user's standpoint, related to above, a centralized subject repository facilitates "one-stop shopping".
(3.) Mirror site creation of repositories may be greatly simplified if all articles are stored in one centralized repository rather than dispersed across numerous small repositories. This is not argument against creating small repositories; it is merely an argument that small repositories are good for certain things and not others. (E.g., they are good for archiving material primarily of interest to a given university or college community.)
(4.) Centralization in major repositories is good not only as a guarantor of archival continuity and permanence, but also from an "image" standpoint: such repositories serve as symbolic bulwarks against unjust pricing schemes of purely profit-motivated entities. Why not just work toward having one subject publisher of journals use one subject centralized repository/database of articles? Such publisher consolidation is very unlikely to happen, given the plethora of publishers out there. While there are arguments for having a consolidated subject repository, it seems unlikely that publisher consolidation would reach the point where one subject related publisher would have an exclusive relationship with just one big subject repository, which it (the publisher) would itself maintain.
20. Might commercial publishers legally challenge implementation of the above model by university publishing efforts?
Not being a lawyer, it is not clear to me how to answer this. Here are some common sense reflections. But they may well reflect my ignorance of the law about unfair competition.
Begin with the following, from SPARC Open Access Newsletter, issue #85 :
" Don't underestimate corporate complaints about unfair competition from publicly-funded OA. Publishers used this argument effectively to kill PubScience in 2002.
http://www.osti.gov/disconps.html
http://www.siia.net/press/releases/111502.pdf "
However, regardless what one thinks of arguments about unfair competition posed by public involvement in publishing, for the following reasons it is hardly the case that subscription subsidized open access model as implemented by private universities would constitute an unfair challenge to business competition. (The question remains about public as opposed to private universities.)
For one thing, much of the material published by commercial entities reflects work done at universities. For another, if arguments were adduced that universities have no right encroaching on commercial terrain, this would furthermore (ridiculously, by consistency) be an argument against university presses or any university-based publishing operation of any kind.
On the other hand, might commercial publishers argue that university presses or traditional university publishing operations are revenue generators and therefore businesses? (I confess ignorance as to how the majority of such university operations are run, that is, for profit, or merely for cost-recovery.) On this basis, might they challenge any contrasting, purely "cost recovery" (that is, non profit motivated) publishing model as implemented by universities? This is not clear, but any arguments they were to adduce would certainly strain credibility at the bit. Certainly the practice of commercial unfair pricing of recent history would undercut these arguments to a significant extent.
Heather Morrison, Project Coordinator of the BC Electronic Library Network, very helpfully suggested terminology for the model described above. She suggested "subscription subsidy model for OA funding", which now --in the form of "subscription subsidized open access", replaces "subscription access" or "subscription funded open access". Thanks to Prof. David Goodman of the Palmer School of Library and Information Science Long Island University for discussing the Medline example and for more generally for his educative contributions to discussions of alternative publishing models. Thanks to individuals in general who took time to respond to aspects of the model detailed above. The journalist Richard Poynder pressed me with questions about aspects of the model. Dana Roth of Caltech and David Stern of Yale stimulated further reflection on some points. Thanks to one of my colleagues for prompting my thinking about the question posed about commercial publisher challenges to university "encroachment". Thanks to James Burr for discussing the analogy between this model and trends in Internet access to information generally. Also, Matthew Cockerill of BioMed Central helped to clarify aspects of some of their journals. Acknowledgments do not imply that the individuals mentioned concur with the content on this webpage.
VIEWS EXPRESSED ON THIS WEBPAGE ARE THOSE OF ITS AUTHOR AND ARE NOT INTENDED TO REPRESENT THOSE OF EMPLOYER.
(610) 758-5003
Last updated 8/2/06