Walk-in access for unaffiliated (walk-in) users in academic libraries is under threat. In North America, this is due to outdated technology as well as licensing changes publishers and vendors impose on libraries in the name of progress. We conducted a survey in autumn 2018 to get a snapshot of the current state of unaffiliated user access at post-secondary libraries across North America. We discuss the current state of access for walk-in users as well as current and future trends in user authentication which threaten this access. We look with a wider lens on open access, community access to paywalled research, and general access to information.

Issues in for-profit scholarly publishing are not new. We are an electronic resources librarian and a systems librarian, and in the course of our research on unaffiliated users and academic libraries, we continued to identify and be stymied by trends in vendor technologies, systems, sales models, and products that caused concern. We hope to share these concerns, summarize the landscape, and also provide some advice on possibilities for collective action as we move forward.

INTRODUCTION Libraries frequently support their open access (OA) fund using money from their collections budget. Interest in assessment of OA funds is arising. Cost per use is a common method to assess library collections expenditures. OA article processing charges (APCs) are a one-time cost for global, perpetual use. Article level metrics provide data on global, cumulative article level usage. This article examines a method and discusses the limitations and implications of using article level metrics to calculate cost per use for OA APCs. METHODS Using different APC models from two publishers, PLOS and BioMed Central, this article presents a cost per use formula for each model. RESULTS The formula for each model is demonstrated with available data. The examples suggest a very low cost per use for OA APCs after only three years. DISCUSSION Several limitations exist to obtaining article level data currently, including the nature of open access and accessibility of the data. OA articles’ usage levels are high and include use from altruistic access. Cost per use comparison with traditional publishing models is possible; however, comparison between different OA expenditures with very low costs per use may not be helpful. CONCLUSION Article level metrics can provide a means to measure cost per use of OA APCs. Libraries need increased access to article level usage data. They will also need to develop new benchmarks and expectations to evaluate APC payments, given higher usage levels for OA articles and considering altruistic access.