A discussion about Open Access

Last week, Richard Poynder (@RickyPo on Twitter) organised an informal get together around open access (OA), that was attended by researchers, publishers and librarians. A broad range of points were discussed, but one aspect caught my attention in particular, which I wanted to write down.

I will start with a comment by Stephen Pinfield (@StephenPinfield on Twitter), who mentioned that many proponents of the green OA route (which I consider myself one of) see it as a way to disseminate research and not as a business model (which the golden OA route, the dominant model in the UK, certainly has become¹). The views of these green OA proponents are often based, rarely explicitly though, on author processing charge (APC), albeit very low ones.

Later, we had a quick survey where all participants shared what they thought a fair APC should cost - this wasn’t meant to be an objective assessment (because a lot of what is said and felt related to OA is very subjective). The range of suggestions ranged from $5 (based on the arXiv model) to $3500. The reason for such a huge variance, and some (arguable very) high APCs comes from cost unrelated to publication. Sudhakaran Prabakaran (@wk181 on Twitter), who has first hand experience with AAAS, explained that these costs include, for example, the news that are published in the journal Science; for eLife, represented by Emily Packer (@PackerLEmily on Twitter), APCs of £2500 (based on this post, down from £3147 projected cost for 2016), include the development of new publishing platforms (that are released as open source software, by the way).

I found it quite revealing; personally, I don’t want to pay for news and views as part of my APCs, but I might consider paying for them independently. Similarly, while I do believe that there is a need to a modern publishing platform, I am not convinced that my APCs should necessarily include such costs; or at least researchers should be able to decide how much they want to contribute to this. And let’s not forget that the money that pays for APCs comes from the funders (my previous sentence is misleading, as researchers don’t pay for APCs themselves).

One comment and suggestion that I found particularly noteworthy was from Rupert Gatti (@rupertgatti on Twitter), co-founder Open Book Publishers. The publication of a hard-copy book at Open Book Publishers ranges between £3500 and $4500 (which is very low compared to the average costs of other mainstream publishers, peaking around 30K for small publisher and increasing for major players). Rupert made an important point in denouncing APCs, in that they explicitly separate research and its publication. This is certainly a major issue in the current climate: researchers don’t pay for APCs; the money comes straight from funders via their university library), and assume everything works behind the scenes, including licensing (and loosing the rights to their outputs). He suggested that if we embedded publishing into research, the cost would become essentially zero (in line with costs of pre-prints). While one could argue that publishing by third-party professionals has some benefits, if we consider that researchers do most of the fundamental work (from doing the research, writing and reviewing the research papers), one wonders if separating research and its dissemination at such a high cost is a reasonable solution.

Finally, we also briefly discussed different publishing models, beyond the green and golden OA models that are currently mainstream.

Flipping, where publishers would flip to a fully OA model and would be paid by bulk deals. This is what I believe the current dominant golden OA model in the UK was meant to lead to.

Institutional publishing, which exists, but is rarely used as a means to comply to the OA mandate.

Funder-based publishing, such as what the Wellcome Trust, the Gates Foundation, … do through the F1000 publishing platform. Ubiquity Press has a similar model, I believe.

And there are probably more that I’m missing… (feel free to comment to expand the list).