Targeting by Numbers. The Uses of Statistics for Monitoring French Welfare Benefit Recipients

Abstract

The targeting of welfare recipients in order to control their situation and their entitlement has increased from the beginning of the 2000s onwards. Data mining has recently been included in the set of techniques used for this purpose, in addition to traditional bureaucratic checks of documents, home inspection visits, and to data crossing. Imported from the private sector, this statistical tool is part and parcel of a “risk management” policy of the family branch of French social security. It has been promoted as the cornerstone of recipients’ monitoring since 2010. Analyzing the use of this method enables us to show the new relationships between statistical instruments, legal norms, and performance indicators which define the administration of the Poor in the neomanagerial era. Thanks to statistical correlations, this tool identifies welfare recipients’ features significantly associated with the highest level of risks of irregularities. Then, scoring algorithms enable local managers to target high-risk populations over which in-depth checks are performed. This has led to positive financial results, but also to an increasing focus of surveillance on the most disadvantaged households. Based on interviews with executives of the National Family Benefits Fund (Caisse nationale des allocations familiales - CNAF) and with local managers, ethnographic observation of street-level bureaucrats’ daily work and quantitative analysis of national and local data, our contribution is twofold: on the use of statistical modeling in welfare policies implementation; on the compounding of control in the contemporary government of the poor.