Data in welfare: biopolitics of algorithmic welfare and the (in)dividual

The home has become a site of interaction with public employment services (PES) in many countries as digital services and creeping digital-by-default models of service delivery were introduced. Further, statistical profiling models are used to categorise unemployed jobseekers in one third of OECD countries (OECD, 2018). Resistance to this datafication of welfare mainly emerged from caseworkers, with several digital-by-default and profiling systems shut down (Riipinen, 2011; Allhutter et al., 2020; Matty, 2013). However, the national employment services strategy in Ireland reinforces a commitment to both digital delivery and profiling (Government of Ireland, 2021). In exploring the datafication of welfare we look to data in welfare. Foucault (1978) debates supervising populations through biopower, using “an entire series of interventions and regulatory controls”. Deleuze (1992) explores the role of computers in societies of control, assigning codes to (in) dividuals to potentially single out “subjects at risk”. This datafication of welfare encompasses the collection of data and the statistical measurement of (unemployed) populations of interest. Using ethnographic fieldwork, interviews and documentary analysis, we investigate how this data is collected, by who and why. Early analysis indicates elements of the datafication of welfare rebranded as a personalised service, backed up by significant state investment in service delivery modernisation.