Science publication should be seen as a public service (just like science itself)

Who should pay for scientific publications: the readers, via subscriptions fees, as has been the universal model until recently? Or the authors/research funders, via APCs (author processing charges)? Often, these are seen as the only two possibilities, but I would like to argue for the superiority of a third option: Scientific journals and books should be published with no cost for either readers or authors by the research institutions themselves.

Research in basic science is not normally carried out by profit-oriented enterprises. Its product (new knowledge) is not readily marketable, but it is universally recognized as being valuable to society, so basic science is normally funded by governments and philanthropists.

But of course, this does not mean that all aspects of science need to be carried out by nonprofit organizations. Many services are provided more efficiently by private companies that compete with other private companies. For example, the people who clean my office are not employees of my research institute. So it could be argued that science publication, too, should be done by profit-oriented publishers, not by publicly funded institutions.

The problem with profit-oriented publishers is that the market doesn’t work: Since scientists have to publish in the most reputable places (journals and imprints) in order to build their careers, they cannot simply switch the publisher if the services are not cost-effective. My institute can easily replace the cleaning company, but it cannot easily cancel journal subscriptions. Journal names and imprints of book publishers are career builders and thus have systemic relevance for science. Where the market does not work, this results in unjustifiable profit margins and inefficiency (as recently described in an impressive way by Stuart Shieber).

Over the last decade, many science administrators and scientists have been arguing for open-access publication, where the readers do not pay anything – they just download the paper (or the book – if they want a paper copy, they can use a print-on-demand service). The idea is that open-access publication will be more efficient (i.e. cheaper overall), because the market will be more transparent. But is this indeed the case?

If journals and books are published by the same companies (the big science publishers, such as Elsevier, Wiley or Taylor & Francis), then instead of charging the research institutions via library subscriptions, they will charge them via APCs (author processing charges). Will this lead to greater market transparency, competition and price fairness? Stuart Shieber thinks so:

“journals compete for authors in a way they don’t for readers, and this competition leads to much greater efficiency. Open-access publishers are highly motivated to provide better services at lower price to compete for authors’ article submissions.”

I don’t share this optimism. Journals with greater prestige would be able to charge higher APCs, and scientists with lower research budgets would end up publishing in the lower-prestige journals. Scientific visibility would be tied even more to research budgets than it is now. In the current system, poorer scientists may not be able to afford subscriptions to a large number of expensive journals, but if they produce excellent output, they can get into the best journals.

Moreover, while it is true that Lingua (an Elsevier journal) and Journal of Linguistics (a CUP journal) are true competitors for author contributions (this was Shieber’s example), this is not the case, for instance, for Linguistics, Folia Linguistica, Theoretical Linguistics, Cognitive Linguistics, and Linguistic Typology: All are published by De Gruyter. And just as publishers want to sell their journal subscriptions in bundles (a practice that Shieber says leads to market dysfunction), they also want to sell their author services in bundles. For example, De Gruyter recently announced that they had signed an agreement with the Max Planck institutes which seems to be aimed at winning Max Planck authors by giving them a discount. So in the future, Max Planck scientists may be restricted to publishing with those publishers that offer a cheap “bundled” price to the Max Planck Society.

The result is the same market dysfunction that we have with the subscription model. The reason is that journal names and imprints are career builders and have systemic function. They cannot be easily replaced. Moreover, because of the importance of the prestige/reputation factor, an economic model that regards science publication exclusively as a service to scientists is insufficient: By publishing my excellent work in a journal or a book imprint, I contribute to maintaining or enhancing the value of the journal name or the imprint. It therefore seems paradoxical that I should pay for a service that I am providing to the publisher.

Thus, I conclude that the best approach for science publication is to consider it a public service that should be paid for by the research institutions directly, without the involvement of profit-oriented companies. Research institutions should fund the publication of scientific results for the same reason that they fund scientific research: In order to make a contribution to the well-being of the society. This does not, of course, mean that there should be no competition. After all, research institutions compete with each other, just not economically. Every research institution wants to be at the top, i.e. it strives to maximize its reputation. Similarly, all publicly funded journals and book series would contribute to their organizations’ prestige. The University of Chicago Press would no longer have any income from its publications, but it would contribute to the University of Chicago’s prestige, just like the university’s research, which doesn’t generate any income either.

It may sound naïve to expect that such a situation could actually come about, but Shieber notes, for example, that two thirds of open-access journals do not charge APCs. It is very likely that these journals are funded in order to enhance their funders’ prestige, precisely in the way I have suggested. This model can be called Platinum Open Access, meaning a subtype of Gold Open Access where the author is not charged.

And at least for the humanities and social sciences, where a lot of research happens outside of special research grants (and significant contributions are made by independent scholars), it’s quite unclear how an APC system would work. In the UK, the government now wants to mandate obligatory (primarily) APC-based open-access publication for all scholars, and humanities researchers are rebelling against this. I think that they rightly fear that their universities will set up committees to preselect the kind of publications that they will support. Martin Eve, founder of the UK-based Open Library of Humanities, writes that they will “waive APCs for those who cannot afford them”, which is nice, but in practice it means that APCs are voluntary contributions, somewhat like the contributions many Americans make to their favourite public radio station. In other words, the Open Library of Humanities will work as a public service, even though it does not have a powerful institution behind it. It is to be hoped that well-meaning governments such as the UK government, who want to promote the transition from subscription-based model to open access, will realize that they will be successful only if they give significant support to Platinum OA.

Thus, it would be great if more powerful institutions get involved in science publication. The recent launch of eLife, supported by three important non-profit research (funding) institutions, points in the right direction.

Postscript: After writing the above, I learned about PeerJ, which uses a very courageous business model, with very low APCs. I’m waiting to see if they have success. In any event, it is clear that the innovation is NOT coming from the established profit-oriented publication giants.