Data solidarity: why sharing is not always caring 

Essay by Barbara Prainsack: “To solve these problems, we need to think about data governance in new ways. It is no longer enough to assume that asking people to consent to how their data is used is sufficient to prevent harm. In our example of telehealth, and in virtually all data-related scandals of the last decade, from Cambridge Analytica to Robodebt, informed consent did not, or could not, have avoided the problem. We all regularly agree to data uses that we know are problematic – not because we do not care about privacy. We agree because this is the only way to get access to benefits, a mortgage, or teachers and health professionals. In a world where face-to-face assessments are unavailable or excessively expensive, opting out of digital practices would no longer be an option (Prainsack, 2017, pp. 126-131; see also Oudshoorn, 2011).

Solidarity-based data governance (in short: data solidarity) can help us to distribute the risks and the benefits of digital practices more equitably. The details of the framework are spelled out in full elsewhere (Prainsack et al., 2022a, b). In short, data solidarity seeks to facilitate data uses that create significant public value, and at the same time prevent and mitigate harm (McMahon et al., 2020). One important step towards both goals is to stop ascribing risks to data types, and to distinguish between different types of data use instead. In some situations, harm can be prevented by making sure that data is not used for harmful purposes, such as online tracking. In other contexts, however, harm prevention can require that we do not collect the data in the first place. Not recording something, making it invisible and uncountable to others, can be the most responsible way to act in some contexts.

This means that recording and sharing data should not become a default. More data is not always better. Instead, policymakers need to consider carefully – in a dialogue with the people and communities that have a stake in it – what should be recorded, where it will be stored and who governs the data once it has been collected – if at all (see also Kukutai and Taylor, 2016)…(More)”.