In a sequel to his Oct 9 blog, Anton Angelo writes on what happened next in their investigation of APCs at the University of Canterbury.

Let’s face it, Article Processing Charges (APCs) are a pain to understand and manage.  APCs devolve the costs of scholarly publication away from from the library, where subscriptions can be neatly reported on, monitored and centralised, to being the responsibility of individual researchers where payments are currently almost entirely untraceable.

We last left gentle readers with our efforts to understand how much the University of Canterbury was paying in APCs (Follow the money: tracking Article Processing Charges at the University of Canterbury).  Our next step was to launch a pilot where we had some central funds to pay APCs, and asked for applications.  This is the story of that pilot, and how it proved useful far beyond simply providing needed cash for researchers to make their work Open Access (OA).

The initial idea for the pilot came as a recommendation after we had analysed the data of our survey on APCs.  There were a number of threads we recognised needed to be addressed:

• The order of magnitude of the problem was in the hundreds of thousands of dollars (and probably higher), an amount of spending by the university that could not be ignored.
• If the institution was supporting Open Access, then we needed to support that practically, whatever our thoughts were about Gold v. Green (or the horrid hybrid). Researchers should publish in the best possible place for their work, and when that was in an OA journal that required APCs, then it is incumbent on us as an institution to help with that.
• Libraries are constantly concerned about maintaining their relevance and launching an initiative kept the library in centre of the scholarly publications process.

The fund itself was set up quickly – $NZ10,000, and a six week application period at the end of 2015.  We knocked up a few web forms, sent out emails to Heads of Departments and advertised in the internal bulletin.  Criteria included; that local scholars were the lead author, journals and publishers were ‘reputable’ and early career researchers would be preferred if the fund was oversubscribed.  A crack committee of interested academics was assembled to assess the applications, chaired by me, as a facilitator.

The first application came in within an hour of the web form going live.

In all, six applications were received, from all parts of the University, totalling a request of $NZ11,000.  We found another thousand dollars so we didn’t have to reject any applications because of money, and went through them one by one.

The applications could not have represented the current state of Gold OA better.  Three of them accounted for 90% of the money, with APCs of $2,500, $3,000 and $4,500 apiece.  The most expensive was a hybrid journal, with a publisher that had no concrete reporting on how APC payments affected our journal subscription rates.

Two of the applicants were for PeerJ author memberships for one article.  The committee exercised itself over these, as the cost was for the submission of the article, not for its publication after it was accepted, as is the PeerJ way.  Though we decided to fund the membership only if the article had been accepted (and let the poor researchers suffer a financial loss as well as the indignity of rejection) the idea of sponsoring researchers to publish – and thereby sponsoring the model and publisher itself, was found to be Very Interesting Indeed.

The last application was the hardest.  From junior faculty, it was a request to publish with a “suspected predatory publisher” according to Jeffrey Beall.  We had included Beall’s list in our criteria of ‘reputability’ as well as h-index, inclusion in the Directory of Open Access Journals (DOAJ), and impact factors.  I had argued against formally including Beall’s list, as I have an issue with black lists being problematic. White lists and selection criteria are harder to manage, but fairer and less liable to bias, but Beall’s list has captured the academic imagination, so it was in.

I was trepidatious to tell the researcher that we had rejected their application and why.  In fact, the conversation went really well, and a teaching moment on the dark side of scholarly publication was not missed.  A new journal was targeted, and if the APC for that is applied for in a new round, it will be about NZD$3,000, rather than the predatory publisher’s $150.

Dealing with hybrid, predatory, ‘straight’ OA, and PeerJ’s memberships, virtually all the OA business models were covered.  A really strong set of examples mean that the pilot APC fund not only met our objectives above, but also let us work in practice with the implications of supporting OA.  We grew a supportive team in our assessment committee, who can take the message back to their communities, and we have successfully placed the library at the heart of the matter, practical, effective and principled.

The recommendation from this exercise is a strong one to continue funding OA APCs from a central source, if it hasn’t been allowed for as a specific line item in a grant application.  As more funders demand public access to research outputs as a result of their philanthropy, it will move from the policy of the university, to a practical necessity to have mechanisms in place to pay APCs.

We will never be in the position where all of the money that goes towards scholarly publication will neatly lie in one budget for journal and book purchases ever again.  This collaborative, but library led, approach was really successful as one of many ways we are going to need to rethink how we pay the costs for new knowledge to be disseminated.  APCs are a real pain, but fronting up and asking for support from your community can ease that, and even help with the communication about OA and changes in scholarly publication in general.

Postscript.

Once all the dust had settled, we asked for invoices so we could pay for the APCs we had funded. There was a little urgency, as we wanted to spend the fund within a particular budgeting period.  Of course, none of the authors were at the stage where they needed to pay for their APC charges.  Publishing is a notoriously slow and involved process, so we should have expected that this part of the process would be delayed. Discussions with colleagues who had run similar programs revealed they had suffered from the same problem. When this process scales up to cover more of our research output there will be a considerable administrative workload as the Library shifts from paying for entire databases worth of articles in one hit, to paying to publish each article from your institution.  Though most publishers offer institutional deals for covering APCs, they are not worthwhile to us from a purely financial perspective.  When we have a larger volume of publications APCs are due for, and once we account for the time of handling each invoice and payment, it may make more sense to do some kind of bulk deal with the bigger players.

 Anton Angelo is Research Data Co-ordinator, University of Canterbury.

Contact: @antlion