Why I think Plan S is not the right solution

There has been a lot of discussion about open science in the last years, and rightfully so. With the ever-increasing importance of science and technology in our everyday lives, and especially the public interest in the science regarding the understanding of the novel coronavirus pandemic, the public should demand transparency and accessibility of its funding agencies and scientists. One proposed solution is the Plan S. Plan S proposes to have funding agencies require within their grant agreements that scientific results achieved due to their funding be published in Open Access journal articles (Open Access means that it is free for anyone to download – no subscription fee). But beyond that, Plan S also demands that the Open Access journals be only Open Access, i.e. the journal cannot give an option to be Open Access or not (“hybrid”), it must publish only Open Access articles. There are many specifics involved, and recently Plan S has proposed a softer set of requirements, but I hope here to discuss some points that I don’t often see.

I am quite skeptical of Plan S for many reasons. Some of these reasons have to do with my own situation and my own biases – I am an early career researcher with very limited funding (actually right now I have no job, but even if/when I have a contract in the near future I will definitely have limited discretionary funds). But before I detail my criticisms, I do not mean to be completely opposed to the idea either. I will detail some recommendations that I have as an alternative because it is clear that if some journals want to charge close to 10k for Open Access, then we need to have more options or at least better common practices.

The reason that Plan S is opposed to hybrid journals is because they believe that hybrid journals do not help transition to more Open Access quickly enough. Basically if you have to pay a journal to be Open Access, but you aren’t required too, then the incentive is only there for certain well-funded labs. But, Plan S policies offer their own incentives that are worse for the research world.

Having paying to publish as the norm creates a perverse market incentive for the publishers. In the current system the journals that have the best work and do the best editing can charge the highest subscription fee. Their incentive therefore is to provide the most value to the reader (not the most value to the author, as all authors can attest). But, if the model shifts such that authors submitting papers pay a fee, then the journals have the added incentive to publish as many papers as possible. This incentive could lead to the peer-review process degrading in favor of quickly publishing articles, well-funded labs salami slicing their work to publish more papers, journals introducing fees to “fast-track” your peer-review, and maybe most importantly, less ability for a reader to discern what articles are worth reading and which journals are reputable sources.

Being required to pay for Open Access is most difficult for the least privileged researchers. This is a commonly made point, and one that speaks to my biases. If the model shifts from institutions paying subscriptions to individual researchers paying publishing fees, then the least privileged researchers are hurt the most. This applies to early-career researchers who haven’t yet secured significant grant funding, researchers from countries with worse science funding infrastructure, and even (yes) independent researchers that are not salaried. For all of these people a 10k fee is impossible, but even the 1-2k fee that is basically standard even at low-level journals is still not accessible. This is a really big deal for researchers themselves of course, but it is bad for society because it would contribute to a less equitable and fair environment for researchers to succeed and a pool of open access paper that do not represent all of the science out there. Maybe some universities or institutes will make it a policy to pay the fees to publish Open Access on behalf of researchers, and this is what I propose is also required when a funding body joins Plan S, but as far as I know this is not anywhere near the norm.

The banning of hybrid journals is just side-stepped by publishers. Publishers are smart and will take advantage of a loophole when they see one. So, if you submit to Journal X (a hybrid journal), your paper is accepted, but you then request Open Access, the publisher transfers your manuscript to Journal X Gold (an open access journal that is not hybrid – a “Gold” Open Access journal). This way you fulfill your requirement under Plan S and the publisher has still attracted you to their journals. There is effectively no difference between this scenario and just publishing Open Access in the hybrid Journal X, but a needless journal has been added to the already diluted and overfull world of publishing with tens of journals for every topic. Why not simply allow hybrid journals?

Mandating strict policies from top-down is against the spirit of academic freedom. These days researchers – even top-level researchers at prestigious institutions – depend almost exclusively on outside funding for their research. There are of course always restrictions on how the funds should be used, and for what, but generally not to the point that the funding bodies dictate the exact content of the work, how it is done, or where it is published. This is not the same academic freedom from decades ago where a university allows a tenured professor to do what they wish, but it is still something. The funding bodies give confidence to the researchers that they know best. Mandating where and how researchers should publish runs contrary to these principles that are already being eroded from all sides.

Having research articles accessible to the public is actually not the most important goal nor the most efficient. The goal of open research is to make the public benefit even more from the research being done at their cost. For me the most important step is not making technical research articles free to read for everyone. These articles are not meant for laypeople to read and are often not even meant for general scientists, but are meant to communicate important results to other specialists in a given field. These specialists are already working for institutions that have subscriptions and can already read the articles. What is very important for making science more open and efficient, however, is to make open source and readily available any and all details possible to replicate or expand upon a work: data, code, methods, etc, etc. In physics and optics at least, this is becoming the norm, and as far as I know is compatible with policies of almost all journals. Detailed supplementary materials (always Open Access even if the article is not) provide minute and boring details which in the past were often left out of articles; codes are more and more often put on github for all to see; there are more options to upload data to safe and reliable repositories, and complex data are more often in standard formats. What do I propose instead? I propose that instead of requiring that an article is open access, require that code, data, etc. is available immediately, and that a detailed summary for laypeople, written by the original authors, is posted publicly alongside the article when it’s published.

The content of articles can usually be found online anyway. With the increased use of the arXiv and bioRxiv preprint servers (and others for social sciences), many authors are uploading a raw form of their article before it is even submitted to a journal, completely free to read and archived long-term. Yes, this often must remain the version before peer review and before being put in a nice format, but all of the info is there, right? It is the info that is important anyway. For articles that make a big splash in the press there are going to be articles in magazines, on phys.org, or on the institute’s website, or on the scientists’ blogs. And lastly, if you are a layperson who just has to read the final published article, there are ways to do that (google it). If all of the information in a scientific article and sometimes almost the exact article itself are already available online, what is the added value of requiring open access for everyone. Of course this is not an argument against Open Access in general, but the point is what is to be gained by mandating Open Access from top-down when so much of the information is already available? And does it overshadow the other negatives above?

I strongly believe in the points above. I also see the point that science should not be constrained by the profit-driven motives and gatekeeping of the traditional subscription-based publishing houses. The current system is by no means perfect (I should make another post why not), but I also don’t want Open Access journals with low standards predating on researchers who just want to have more papers out there to get their next promotion. There are already many examples of a middle ground. OSA Publishing, publishing on optics and photonics for decades, has multiple Open Access journals (some of which were not always that way, but now have their entire library accessible) and is the publishing arm of a non-profit scholarly society. This seems to be a great approach, but they are understandably against Plan S. The journal Physical Review Accelerators & Beams – a journal mostly for research on particle accelerators – is fully Open Access due to contributions from research institutes and private companies (with no strings attached on the journal).

The stresses on researchers young and old to publish or perish is becoming only worse as the number of PhDs continues to increase without being met with new jobs, third-party funding becomes the norm, and universities rely upon incomplete metrics to make hiring and promotion decisions. Something must be done about this for sure, and I hope it will be, but let’s not make it worse by making it more expensive for researchers to publish and dictating where they can publish, when there are many other more efficient ways to make science more open, collaborative, and accessible. Plan S is doing a good job of sparking a debate and pushing researchers and publishers alike to have hard discussions and change common practices – for the better – but the specifics of Plan S are not the right solution.