Transformative agreements are an increasingly common way for universities and consortia to shift publisher business models towards open access. They do this through a prearranged payment that allows institutions to access subscription content while allowing future research to published in an openly accessible form. These deals are a way for publishers to continue to receive subscription income and boast about their open access content, while universities value them as a cost-neutral strategy for transitioning away from subscriptions towards open access (read Lisa Hinchliffe’s primer for an excellent summary of transformative agreements)

Such agreements are being announced by universities and national negotiating bodies on a weekly basis. This week Elsevier announced a pilot agreement with the University of Florida, while Taylor & Francis promoted its new deal with a Finnish library consortium. In recent weeks we’ve also seen deals announced between Springer-Nature and the University of California, and Elsevier and a consortium of Dutch universities. It has been pretty clear for some time that transformative agreements are going to be a dominant model for the transition to open access, particularly for the larger commercial publishers who are quickly learning how to work these deals to their advantage.

There are a number of critics of transformative agreements. Many commentators argue that these deals have a tendency to benefit the publisher rather than the university, as Sicco de Knecht argues with respect to the Elsevier-Netherlands deal. Similarly, Dave Ghamandi describes these deals as a ‘con’ that does nothing to shift power away from the ‘monopolists’. I’m embedding Dave’s Twitter thread below because it is an excellent and impassioned argument about commercial ownership of scholarly communication and the way that transformative agreements merely reinforce rather than counter this problem.

Yet this post is not about the shortcomings of transformative agreements per se, but rather the conditions attached to them. It is worth considering that these arrangements are not merely about access to content and that other factors, described by Hinchliffe as costs, copyright, transparency and transition periods, are all up for grabs in the negotiations. Libraries want these agreements to be cost-neutral, or ideally cheaper than current subscription deals, and there is an important eye to who owns article copyright and what the pace of transition to open access will be. These conditions all affect the overall deal.

One feature of transformative agreements that keeps cropping up in press releases is researcher data. Many of the press releases announcing such deals mention pilot studies to allow universities to learn more researcher behaviour. For example, the University of California-Springer deal will be launching a project ‘to explore the impact and scope of open science content which will provide insight and learnings that can be exported elsewhere in the US, and beyond’. Similarly, the recent Elsevier-University of Florida deal will use user data to ‘to better understand researcher behavior and to learn from the resulting data’. Although it is rarely possible to analyse these clauses in any detail, because they are only pilot studies and much of the information is not yet public, it is clear that transformative agreements imply a level of datafication that was not part of previous subscription deals.

Datafication simply means turning something into data, which in this case is researcher behaviour. As Mejias and Couldry write, datafication entails the ‘quantification of human life through digital information, very often for economic value’. It is a way of rendering human interaction machine-readable in a way that makes it available for further analysis (or for sale to those who want to analyse it). Though there are many uses for datafication that aren’t all bad, it also has an ethical component that is particularly associated with surveillance and the darker side of our postdigital world (surveillance capitalism, Cambridge Analytica, and so on). Commercial publishers are therefore using transformative agreements to make researcher data available for their own use and to develop products and services to ultimately sell insights back to universities.

Although it is not a new observation that publishers hope to capture researcher data in this way (Sarah Lamdan’s piece on librarianship and ICE surveillance is a must-read, for example, but many others have written on it), I feel that little attention is being paid to the datafication inherent within transformative agreements. As many open access advocates (rightly) critique the regressive elements of such agreements, it is important to also keep in mind that publishers are also on a mission to create the conditions for extracting researcher data wherever possible. This is to say that the transition to open access is not about mere access to research but rather the broader cultures of knowledge production and the ways of interacting with a publishing industry that is rapidly turning itself into an ‘information analytics’ industry.

It is tempting to frame datafication as a Trojan horse that negotiators are unaware of when making these agreements, but this would be a mistake for two reasons. Firstly, the turn to datafication in higher education is welcomed by universities who are looking for better ways to assess researcher performance. Much like many universities are mandating ORCIDs to understand your behaviour, so too will they welcome the kinds of insights generated through such analytics — administrators are looking for a complete picture of performance across the university. Secondly, and perhaps more cynically, universities understand that they are negotiating access to publications not just using money but using their researchers as well. Researcher data has an economic value that university negotiators can exchange to keep costs lower (at no cost to the institutions themselves). This will be all the more important in the post-pandemic economy in which budgets are tightened and researcher data may increasingly represent a bargaining chip for cash-strapped universities.

The datafication embedded within transformative agreements is worrying not just because of the increased surveillance it will entail, it also illustrates more general misdirection of the transition to open access and the potential danger of universities to use researcher data as part of negotiations. Open access was initially premised on the idea that publishers are extracting from the free labour of academics and privatising the gains through closed-access publications. But through transformative agreements, publishers are still parasitic on this labour in addition to their new strategies of extractivism.