Guest blog post by Jan Erik Frantsvåg, Open Access Adviser at University Library of Tromsø and chairman of the board of SPARC Europe
In discussions, I often hear the cry «let us get rid of the publishers – let us do this ourselves». I understand why this is said – e.g. the extreme profit levels of some major publishers – still, I am sceptical.
It’s obvious that we need to do something about how the market works, and who has ownership of and control over content. Giving away content and buying it back, ridding ourselves of potential readers and creating superprofits for the publishers in the process, is something we should stop doing.
But I think that getting rid of (commercial) publishers as such is a bad idea. The debate should be about what role(s) publishers should have, what business model(s) they should use, and who they should be.
Should we – institutions, small or large – see ourselves as future publishers? Should we take over the major commercial institutions dominating much of the market and all of the debate?
It is possible, but not on the scale I hear many of the critics espouse. I would rather spend USD 5000 per article with Elsevier or Springer for an OA article, than burning off much more money than that with a small, inefficient not-for-profit publisher – unless this is for a transitional period.
Publishing has economies of scale, this means any new publisher must be aiming to publish much more than any single institution’s output in any given field. Numbers from eLife – a not-for-profit publisher – suggests that a quality journal will need to have high numbers of articles to get costs down to a level below what commercial publishers currently charge to make an article OA.
Hidden costs with “free to publish” OA journals
The fact is that costs are often overlooked with “free to publish” OA journals. Costs of such journals often are a) hidden in the accounts of an institution, becoming invisible; or b) kept down by keeping the technical quality low, with the consequences that it might have in the longer run. A practical example of a combination of a) and b) is one I often see.
The professor as typesetter
The cost of using the professor is invisible, as (s)he just spends time typesetting that could have been spent doing research – or family life. The quality of typesetting is – generally – much lower than what a professional typesetter could have delivered. The professor spends more time performing low quality work than a professional typesetter doing higher quality work, and – hopefully – the professor earns more per hour than a typesetter.
In sum, we spend more money this way
It is just conveniently out of sight for us. There are many facets of journal publishing that works the same way. The result is that small, stand-alone journal publishing is costly publishing, with content losing in the fight for visibility because such journals have little room for funding and developing publishing competence.
We cannot, by definition, get rid of the publishers – at least not if we intend to be published.
I repeat – let us be clear that the debate has to be about what role(s) publishers should have, what business model(s) they should have, and who they should be. Not how we get rid of them.
In my next guest blogpost here I will discuss further the need of sustainable APC models.
Stockholm University Press Blog 
I wonder how SciELO, Hindawi, ScienceOpen, F1000 Research etc. all manage to publish at less than 10% of the average subscription article? SciELO is doing so with ~1000 different journals and for more than 15 years already . If the data is anything to go by, their quality (measured as methodological reliability) is actually higher than that of more established journals (unless you have other data than we?):
http://journal.frontiersin.org/article/10.3389/fnhum.2013.00291/full
Hence, the data suggest we can save more than 90% of the money we are currently spending on subscriptions and combat the replication crisis by actually getting rid of publishers.
Getting our content back into our control has never been more obvious as now, when publishers hold our content hostage to demand ransom in exchange for content-mining access to it.
In summary, the underlying root evil of most scholarly problems of today is the obscenely expensive anachronism of a public good in private hands.
P.S.: If one were to define publishers so broadly as to include software companies, then the company that provides the software that powers stockholmuniversitypress.se is the publisher and not Stockholm University Press, oddly enough.
In my understanding of the word publisher, it is the person or organization making content available for the public. So if we want content made available, there must be a publisher of some sort. It could be a university, or a commercial publisher – but we cannot get rid of the function of publisher.
Getting control over the content is the crucial point. Well-constructed OA, using CC licenses, allows science to control content – which is a public good – whether the publisher be a commercial entity or not.
I’ve read somewhere (don’t ask for a reference!) that Hindawi’s profit margin actually is higher than that of Elsevier. When you compare this to the fact that Hindawi’s average APC is lower than Elsevier’s profit per article, it is obvious that publishing can be accomplished at cost levels radicallly different from what could be inferred from today’s subscription and APC levels at major publishers.
Converting to OA will make content a true public good, and will enable new entrants to the market, thus creating conditions for much lower costs. It might be that universities could help create such new entrants, but they should be aware you cannot do this on anything but a large scale.