Will Open Access inhibit innovation?

Last week was Open Access Week. At the risk of sounding like a stick-in-the-mud, let me play devil’s advocate to the blogosphere’s near-universal celebration of Open Access (abbreviated, OA). Thus: I don’t think most OA advocates have thought deeply enough about long-term implications.

First, though, what is Open Access? 

OA is a publication model where scholars (or their subsidizers) foot the bill and readers enjoy studies free of charge. Anyone can log on and read an OA article with nary a registration or fee.  OA marks a radical change from the traditional model, where most costs are paid through subscriptions, usually on the part of universities, corporations, or individuals. In essence, with OA we exchange a demand-driven market for a supply-driven system.

As a frequent consumer of scientific articles, I admit a fondness for OA. I love being able to read an article anywhere. No logins, no payment for pdf. It’s great.

I am not the only one, either. OA has become increasingly viable as publication costs decline and as scholars depend more on publication record to secure positions and promotion. In the present environment, where publications are the prime metric by which careers are judged and where university openings are typically met by hundreds of candidates, academics are willing to pay significant sums to get their work noticed.

So too are corporations, whose product tests are often reported in the same outlets. Here should be the first hint of a problem, if you’ve not spotted it yet.

Imagine a journal with a slim bottom line. Under the subscription system, a journal could attempt to improve their books by courting content more suited to what subscribers were willing to pay for. And most subscribers want high-quality research. Now imagine the decision making process when readers don’t matter. At least, not financially. What is to stop a struggling journal from accepting flawed but well-paying articles? Where is the incentive for scientific integrity?

It can be argued of course that there is plenty of flawed work and even fraud under the traditional system. True enough. But OA is not only unlikely to fix the problem, but it may add perverse incentives for worsening it.

OA advocates present their model as integral to scientific innovation.  To an extent they have a point: as more readers gain access to scholarly articles, ideas will flow more freely and efficiently. But the reality will be more complex.

Consider the following simple deduction. Under a scholar-pays scheme, the well-funded scholar will publish more. (Or at least, since OA articles may be cited more frequently than subscription articles, those with more money will be cited more than those without.)  So if you are one of those tweed-wearing, hand-wringing, dusty old traditionalists who has lamented the decline of the humanities and other less-funded fields as universities increasingly shower favor on their glitzy biomed departments, then OA could make the problem worse.  Much worse. More publications by wealthier departments beget more grants, beget more publications, and beget more positions. The status quo will prevail, and OA may become yet another mechanism by which the powerful retain their hegemony.

Fundamentally then, Open Access may provide for the entrenchment of the already well-funded. Over the long term, I worry that short term gains to innovation will be offset by more substantial hurdles to those scholars with more radically novel ideas.

In spite of what I write here, I am not opposed to Open Access. OA will bring plenty of benefits, especially to lower-income institutions that cannot afford the often astronomical subscription fees. It’s just that I find some of the blogospheric OA cheerleading a little bit naive. Changing the publication system will produce new winners and new losers. They won’t always be who we think they’ll be. Shouldn’t we exercise just a bit more caution?


  1. #1 Scott
    October 24, 2009

    This post echoes a lot of my own thoughts on this.

    It seems an inherently flawed system to essentially buy your publications. Some of the newer OA journals already have author instructions that could be paraphrased along the lines of “if you pay for it, we will publish it”. They have astonishingly high publications costs too.

    Think about the grad student doing awesome work on TA funding (of which there are many), and/or an advisor that doesn’t have the grant money (or won’t spare any for an unrelated project – also very common). It is easy to see that a purely OA system would only favor the well-established and kill a lot of great work, especially from younger researchers.

  2. #2 Roberto Keller
    October 25, 2009

    In essence, with OA we exchange a demand-driven market for a supply-driven system.

    From the standpoint of who pays the costs of publication, this characterization of the OA system, albeit simplistic, has some truth to it. But publication cost is not what drives the scientific literature market. It is a combination of the the journal’s reputation, the scientific quality of the papers and authors appearing on the journal’s pages, and the editorial machinery behind the scenes (good editors, good reviewers, apropiate publication times, etc). And yes, unfortunately, impact factors too.

    You will have journals that publish any fabrication sent to them as long as the authors pay the bill (there are already various cases of such journals). But if science is good at something it is at self-correction. Once these bogus papers are out, they are quickly scrutinized and signaled for what they are. If a journal shows a record of fraud publishing scientists turn away from it, both from submitting as from consuming papers. And no journal recovers from a tarnish reputation, certainly no scientist does.

    I think the benefits of OA for low-funded scientists will vastly outweight the disadvantages. After all, if a taxonomist working in a small university in Latin America can’t access the relevant scientific literature, getting the extra funding for publication cost is the least of her problems.

  3. #3 Marcio Pie
    October 25, 2009

    But there’s a crucial asymmetry, Roberto.

    OA tend to be more cited by researchers in developing countries, but their publications in those same journals is disproportionately smaller. (I think R. Gadagkar indicated this in a recent paper).

    Perhaps we need a combination of OA and a clear commitment of the journal for being non-profit, with open access also to information on publication costs, so that they would be the least possible.

  4. #4 Anders
    October 25, 2009

    Did you see this:

    I think it addresses many of your points.



  5. #5 janne
    October 25, 2009

    This came up a couple of years ago in a ScienceBlogs blog; it turns out that closed journals on average have higher publishing fees than Open Access ones. To the extent that publishing fees are a problem it’s certainly not an Open Access specific one.

  6. #6 myrmecos
    October 25, 2009

    While that may be true now, I suspect it’s confounded with the delivery medium. OA journals are more likely to be online only (less expensive) and subscriber journals are more likely to be printed (more expensive). Does the relationship hold up when controlling for the publication medium?

  7. #7 myrmecos
    October 25, 2009

    Thanks Anders! I wish I’d seen that (and this: ) before I wrote this post.

  8. #8 myrmecos
    October 25, 2009

    Do you think this dynamic might cause the system to gravitate towards a small number of high-quality OA journals, a big subscriber-based middle, and then a mass of OA spam? In other words, something similar to our current television system with its free major networks and subscriber cable (NBC = PLoS; THC = Zootaxa; HBO = Nature; QVC = I hate to think about it).

    Incidentally, I’ve never had a problem getting closed journal articles without a subscription- authors unfailingly send along copies on request…

  9. #9 ihateaphids
    October 25, 2009

    It’s not medium specific, because most European journals don’t have publication charges, even those that are printed, while US based publications have very high page charges.

    Also, I didn’t pay a cent for my PLOS One paper. All the PLOS journals at least will waive the fees if you ask. I just told them the paper I wrote was developed after funding for the project was over and asked for a waiver and never received a bill.

  10. #10 Donat
    October 26, 2009

    The impact from OA on scientific innovation can not be considered just from the point of view of publishing. Access to publication is a strategic asset that already exists, and contributed to a huge North-South divide, as well as within wealthy cities like New York, since not all universities and other institutions can subscribe to the disproportationally rapidly growing costs of e-journals.

    So, working as taxonomist, let’s say at the AMNH, has a huge advantage over most of the world’s places outside the few large centers. So, there is already a disequilibrium between the haves and have-nots (or not so much).

    OA is not author pays, but institution pays, as much as the author has a library that pays for the subscription and not the single user. Institution can be your own institution or your funding agency that increasingly pays for publishing.

    OA is not just upfront payment, considered the Gold Road. There is a much more widespread Green Road, that is self archiving. Figures show that 92% or higher of all the journals allow depositing your own publication on an institutional archive, be it as the pre- or postprint. Even Nature allows doing this. This is what the NIH requires from all fundees now, that they deposit their publication in MedLine, and this is what the many institutions of the Berlin declaration require.

    The value of OA is not per se a free pdf. The real value is that it allows machines, in simple ways google, in more complex ways semantic tools, to do data mining or harvesting (search eg for Colombia in to get taxa described from Colombia, and imagine this little sample includes all the taxonomic publications).

    This real useage of the Internet works only with OA.

    Finally, if we adjust our publication to this latter potential and move into semantically enhanced publications in which links to the underlying data is provided, all of us, and well beyond the small circle of taxonomists, will have a huge benefit.

    The fear that OA will hamper innovation is from this perspective not justified.

  11. #11 myrmecos
    October 26, 2009

    Thanks, Donat. You’re certainly right about the value of OA from an automated archiving standpoint (looking at antbase, for example!)

    I’m having a hard time making sense of your economic argument re: institutional payment. If the growing costs of journal subscriptions can be weathered only by the wealthier northern institutions, wouldn’t that same problem carry over into the institution-pays model? That is, how would a South American museum that can’t afford subscriptions be able to afford publication costs? Before, they could publish articles but not read them; now they can read them but not publish them. I know that many OA journals give waivers for developing countries, but that still strikes me as being a temporary fix. Ultimately, the money and the power will still reside in the north.

    My concern is based on my (admittedly Econ 101 level) thinking that altering the structure of supply and demand will have impacts on the nature and quality of research dissemination. The OA model simply has too much in common structurally with the payment model for marketing and advertising to sit well with me. This is purely a theoretical issue, mind you, I haven’t seen any empirical evidence that this will become a problem.

    I do wonder if we’ll gradually see a phase-out of journals as we know them, to be replaced by a more author- or institution-based archiving. We already see something like this with many scholars serving pdfs from their own servers. Journals would morph slowly into certification bodies that provide structure for editing and peer-review, but the actual distribution would be carried out by the researchers themselves.

  12. #12 Sarah Shreeves
    October 26, 2009

    I was glad to see the idea of archiving come up as there is certainly more than one way to provide open access to research. Underlying much of the discussion about open access is the issue of control – who controls or owns your research? All too often, I think, we sign away our rights to publishers who then can exert control over that material. If you’ve signed away all of your copyrights and the publisher hasn’t granted you back the right to distribute it, you actually can’t legally put it on your website or into a repository.

    Publication agreements are contracts and can be negotiated. I think it comes back to really thinking about what we want to be able to do with our own research – put it in a repository or on our own website? Use it in our teaching? Reuse parts of it – like charts or graphs – in other papers? Make sure you negotiate to retain those rights. (As an aside, part of what makes publishing in OA journals appealing is that generally they do allow you to retain full copyright in your work or request you publish it under a Creative Commons license (though I recall that has posed issues for use of your photography, Alex)).

    In full disclosure, I manage the institutional repository at the University of Illinois, IDEALS ( which allows researchers to make their published research openly available IF they have the rights to do so. I also lead much of the scholarly communications efforts in the Library here. I certainly have an interest in seeing that we as researchers (and I try to do this when I’m publishing) think critically about the agreements we’re making with all publishers.

  13. #13 Donat
    October 26, 2009

    If the growing costs of journal subscriptions can be weathered only by the wealthier northern institutions, wouldn’t that same problem carry over into the institution-pays model?

    There are few thoughts on that.

    1. OA is not only author pay (Gold Road) on what you put an emphasis. In fact, the so called Green Road to OA has nothing to do with author/institution pay.

    2. There are plenty of journal that are open access that one does not have to pay for it, neither as author nor as reader. Especially journal from the developing world seem to to along this road. But also large institutions, such as the AMNH.

    3. OA is not just the few expensive high profile journals such as BMC or PLoS, or some of the hybrid journals from Springer and others.

    4. There is a very good compilation of OA by Peter Suber, well worth consulting (, and the advocate of the Green Road,

    Stevan Harnard (

    5. OA has nothing to do with changes in peer review, quality of the journal.

    6. OA finally is challenging inhouse and other mainly small publishers (thousands of them exist for taxonomy only) regarding their business model. What are the real figure to publish, and what are the hidden inhouse costs, such as shipping, accounting etc. that often are done by other departments in a museum, for example.

    Facing out of journals.

    This seems to be not really going to happen soon. Most of the efforts in this direction are failing and people are worried to contribute to wiki still publications, such as has been offered for a while be Nature, BMC and possibly PLoS. Also, the very objection in our community against an ammendation in the ICZN for online publication speaks for itself.

    What is happening is that more and more dedicated websites show up, like Atlas of Living Australia that are semioffical in a sense that they function also as governmental references for species and seem to include nomenclatorial changes that are not ruled by the Code, such as subjective synonyms. At some point, such sites might replace publications, but when and how this is going to happen is not clear at all.

    Sarah makes some very important points.

    1. Think twice whether you sign ceasing agreement with publishers. Talk to them or just don’t allow them to get the entire copyright.

    2. What is copyright anyway? Following our reading, all the treatment or descriptions are in public domain because they do not fall under copyright (for more detail on this see

    3. Submit all your publications to a local self archive, or, if there is none, talk to your institution to make this happen. This is one way to provide access.

  14. #14 Dave
    October 27, 2009

    This is a fascinating discussion for a number of reasons, most not directly related to the topic Myrmecos posed. One thing in particular that struck me is the general trend of treating the costs of the scholarly publication system simplistically. For example, Myrmecos states “OA is a publication model where scholars (or their subsidizers) foot the bill and readers enjoy studies free of charge.” Well, if posted on a cathedral door or telephone pole, that may be true, but internet access is not free (not even in universities).

    Correct me if I’m wrong, but Myrmecos’ model of OA seems to be the only mention (and a tangential one at that) of pre-submission costs – the costs incurred by the tax-paying public (through their institutions) that pay to support most scholarly authors. The models (Myrmecos, the strange blog posting that Anders linked, and the interesting Scholarly Kitchen critic that Myrmecos linked) seem to start with the submission to the publisher, as if the works to be published have spontaneously generated. In fact, the pre-submission costs represent a public subsidy to publishers.

    As to what I thought was the main point (Myrmecos: “Fundamentally then, Open Access may provide for the entrenchment of the already well-funded”) – well isn’t that the St Mark’s Effect – or is it the St Matthew’s Effect? Google tells me it is both, and that even Luke’s Gospel has the same message: “I tell you that to everyone who has, more will be given, but as for the one who has nothing, even what he has will be taken away” (Luke 19:26). I submit that the current scholarly publication system runs exactly according to this model and Open Access will prevail only if it increases the ‘them that has gets’ factor.

    For those who are more interested in their science than in academic advancement, there is a way around the pitfalls of ‘prestige publication’ and journals-for-profit publication in general, and that is a return to real ‘vanity publication’. If you think about it, most scientific journals used to be vanity publications – but the vanity of societies, rather than of individuals. That has become increasingly less true as publishers gobble-up the journals of small societies and the larger societies become publishing empires.

    An author and their institution can eschew professional publishers and bear the (not very high) post-submission costs of electronic publication themselves (sitting on the tax payer’s back of course). For example, I currently have a large guide to the identification, distribution, and ecology of a group of arthropods hosted on my institution’s website. The 300+ page and growing pdf can be downloaded by anyone with a fast internet connection. Moreover, it is a work-in-progress and I have been soliciting corrections and improvements and updating the tome on a regular basis. In essence, this is a self-published book, and it goes well beyond the image libraries, interactive keys, ToL pages, etc. that represent my previous “open access” web-only publications.

    Seemingly without noticing it, we have let scientific publication become the copyrighted property of corporations. In this regard, I think Open Access would be a boon if it freed up data and ideas from increasingly restrictive copyright laws. The alternative – a return to vanity publication by scientific societies – may not be generally practical. Many if not most societies seem to be in financial straits. But if the costs of paper publication can be avoided, then publishing outside of journals is an alternative worth considering. It is true that you will get no direct academic credit, but in my experience you will make many students and colleagues happy.

  15. #15 myrmecos
    October 27, 2009

    Thanks for the comments, Dave.

    The first draft of this post actually had a paragraph about the public support of science and the responsibility of the publishing and science establishment to the taxpayer. I removed it for brevity, though.

    The public financial support for scientific research should provide an imperative for open access- I believe this is often mentioned in arguments supporting OA. Conversely, though, if scientists engage in an expensive, career-fueled publication race to disseminate their results in the more expensive OA journals, is it really fair to bill the taxpayer when the same results might be made publicly available in a less prestigious venue?

  16. #16 myrmecos
    October 27, 2009

    Thanks for the insights, Donat.

    Do you think copyright- at least for taxonomic purposes- is worth addressing in the ICZN code?

  17. #17 Donat
    October 27, 2009

    The problem would be to a large degree resolved if we would require that new descriptions ought be open access and part of the registration process at Zoobank. If would also be resolved if we would assure that all the images we take and publish are in archives such as morphbank to which a link could be provided and we control the access.

    As a community, we can not change copyright legislation, unless we build such a momentum as has happened in the biomedical NIH community. But we can introduce best practices, such as above.

    The latter will become more feasible if we would have shared bibliographic repositories (eg citebank that might come up in the context of the Biodiversity Heritag Library), shared archives for images or ontologies to which our terminology maps. With that, treatments could become more and more independent of the surrounding front and backend of a publication.

    This, if it would materialize from within our communities would at some point become part of our science infrastructure, similar to what happened to GenBank – but only if we make this initial effort…