Data-driven governance and performances of accountability: critical reflections from US agri-environmental policy

Public institutions have increasingly responded to calls for more accountability by promoting ideas of data-driven governance. As this focus on using data tools to strengthen governance intensifies, it is important to examine how the tools that underlie such claims are made. In the context of US agri-environmental policy, policy leadership has spoken the language of data-driven decision-making for over thirty years, primarily in response to accountability demands. However, the reliance on metrics and calculation is most explicit in the newest initiative called the Conservation Stewardship Program (CSP). Unlike other initiatives, the CSP used a single environmental scoring and decision-making algorithm to allocate approximately US$3.8 billion to 68,000 farmers. Applying a Social Construction of Technology approach to the development of the tool shows that different social groups - political aides, program officials, and subject-matter specialists - interpreted data-driven governance differently. Tool design emerges not despite but through the contestations over the purpose and practicality of doing data-driven governance. Social groups involved in tool development are themselves embedded in different accountability structures that shape the rationalities they marshal when negotiating practical decisions about tool design. The final tool used in the CSP was technically contested and the tool development process was poorly documented, yet it had an important legitimacy function: it deflected accountability pressures for more rationalization in public spending without causing major policy disruption. Centering analysis on the negotiations entailed in making the tools that come to underlie claims of data-driven governance can explain how and to what extent a push for more data in public policy can strengthen (and weaken) accountability relations.