Paper long abstract:

Big data in biomedical research are neither ready-made, nor readily accessible: and are not simply 'big'. The sociotechnical apparatus governing sharing of these data is the neoteric 'data access committee' (DAC). Still emergent, DACs' remit, composition and guidance vary widely in governing access to study data and/or samples. DACs are a central infrastructure in the international data economy, yet little is understood of their practices, networks and configurations. We conducted ethnographic research in one DAC responsible for access to data and samples in four longitudinal cohort studies.

DAC member's used a range of rhetorical strategies to warrant access/non-access to 1) potentially controversial research and to 2) finite resources. Decisions involved negotiating and traversing: multiple relations with funders, research users, study guardians and policy makers; governance oversight and ethical prohibitions; and, technical and administrative structures and strictures. The paper describes these practices, and the epistemic and non-epistemic values and constructs of legitimacy, accountability and responsibility inhering to them.

In this setting responsibility was a contested terrain where hybrid subsidiary/sovereign modes predominated and formal mechanisms of governance largely comprised delegated (albeit perpetually challenged!) authority. Absent human and non-human actors (study participants, 'the public', mass media, the studies, 'the science') remained present in data access decision making, providing a touchstone for communal, social, institutional, as well as personal modes of responsibility. The undoubtedly asymmetric relations of power/responsibility were not simply or uniformly monolithic. In concluding, we offer a conceptualisation of responsible research and innovation which draws on these findings.