The last week has featured dueling tweetstorms about the article processing charges (APCs) assessed by leading open access publisher PLOS.
First up was population geneticist Andrew Kern, who argues that the APCs charged by PLOS ($2250 per article for him most recently) actually props up lavish executive salaries rather than furthering open access. Slash those salaries and severely reduce APCs, says Kern. In response more researchers would pay those charges so their work can be open access, even when they have to pay them out-of-pocket. In other words, "chop from the top."
Michael Eisen, biologist and one of the founders of PLOS, responded quickly. Eisen agrees that APCs are too high and wishes they were lower. But he also argues that the PLOS board has a fiduciary responsibility to ensure a healthy operating margin for the journals, to ensure that PLOS survives and can continue to build infrastructure to support open science (which is more than open access). The CEO salaries are the going rate, says Eisen. He wishes they were lower but if PLOS went below market they could not attract top talent. Publishing is a lot more expensive than people realize, notes Eisen, in perhaps the one instance in which he agrees with the claims of leading subscription-based publishers.
Eisen also points out the rich irony: when PLOS began everyone said open access was a crazy idea, with no way to make money. Now PLOS is dinged for making too much money. Damned if you do, damned if you don't.
When open access entered common parlance in 2003 -- thanks in large part to PLOS -- I echoed many librarian colleagues in hoping that this development would lead to much reduced prices for accessing the scholarly literature. Obviously the costs ended completely for end users. But hopefully they were also going to drop dramatically for institutions such as academic libraries, which have the mandate to collect and preserve this literature.
This was our naive hope back then. More sagacious observers could see that open access would not necessarily be a cost saver for academic libraries -- rather, the costs would shift from paying for subscriptions and licenses to paying for article processing charges. Indeed, for very prolific research institutions costs could rise dramatically in an APC regime. Because for all the good open access achieves, it does not end the "publish or perish" mindset. And if anyone else pays APCs, the authors of scholarly works remain shielded from the actual costs of production.
So over the years I abandoned the hope that "gold" open access would save money for institutions. This was never the goal, ultimately -- the goal was much better access to the literature, with fewer restrictions on re-use and re-purposing that content. If this cost the same amount as the subscription model -- even if it cost more -- there was a salutary and undeniable public good. Not only would everyone have access to the scholarly record as it is being created. Librarians would no longer have to erect elaborate authentication systems and lock-out mechanisms to keep the unanointed away from our networks.
And yet. And yet....
In all these years of advocating for open access, I have never had a good explanation for why APC charges vary as widely as they do. Nor do I actually know what the money goes for, down at the article-by-article level.
This is similar to receiving an undifferentiated bill for services at the hospital. You see the grand total but not the line items that produce that total. As Steven Brill reminded us with Bitter Pill (2013), this lack of transparency leads to grossly distorted health care prices in which the cost of service bears little relation to the services provided.
It is hard not to believe that the same thing occurs with APC amounts.
A few authors are motivated by altruism, many others are motivated the fact that their funders insist on open access (often with embargos at the moment, but that feels like a transitional stage toward immediate open access.) Those funders generally pay very handsome APC charges, as a recent Harvard Office of Scholarly Communication report demonstrates. In the context of a funder's overall grant portfolio, even a very high APC is pocket change. Motivated authors are connected to flush funders, who pay non-itemized bills for article processing charges.
In such conditions there must be some level of APC inflation.
This does not mean PLOS is nefarious, or is not genuinely committed to open access. But as a business entity too, the perceived need to ratchet up APCS beyond what would strictly be necessary (even granting the need for healthy operating margins) must be irresistible.
Or to put it another way: If Ubiquity Press can charge $400, and many journals on the DOAJ charge nothing, why was Andrew Kern's bill $2250?