Same, same, but different! Qualitative evidence on how algorithmic selection applications govern different life domains

2020

Noemi Festic

Regulation & Governance [Online first]. doi: 10.1111/rego.12333

The term algorithmic governance describes institutional steering effects of algorithmic‐selection applications that increasingly pervade all domains of everyday life. Empirical evidence on algorithmic governance is lacking and mostly limited to information services. This article compares the significance of algorithmic governance – measured by use, subjective significance, awareness, risk awareness, and coping practices – for four pivotal life domains (information, recreation, commercial transactions, and socializing). Drawing on qualitative, semi‐structured interviews with Internet users, this article reveals important nuances in how differently users engage with algorithmic‐selection applications across life domains and functional types like search or recommendation. While awareness of algorithmic selection and related risks is comparatively higher for information services, the findings reveal a significant lack of knowledge for algorithmic selection in other life domains and for specific algorithmic modes of operation. This article provides input for an evidence‐based development of suitable regulation of algorithmic‐selection applications, taking everyday practices of their users into account. http://dx.doi.org/10.1111/rego.12333