The street-level realities of data practices in homeless services provision

Quantification and standardization of concepts like risk and vulnerability are increasingly being used in high-stakes, client-facing social services, also presenting the potential for data-driven tools for decision-making in this context. These trends necessitate an understanding of the role of quantitative data in the work of street-level decision-makers in social services. We present a qualitative study of existing data practices and perceptions of potential data-driven tools in housing allocation, engaging the perspective of service providers and policymakers in homeless services in a large urban county in the United States. Our findings highlight participants’ concerns around centering clients’ choices and ensuring integrity in a resource-constrained, high-stakes context. We also highlight differences between the perspectives of policymakers and service providers on standardization and fairness in the decision-making process. We discuss how use of and policies around data in social services need to consider the importance of the relationships that client-facing service providers have with other workers in the organization, with their work, and with clients. © 2019 Association for Computing Machinery.