An instrument constituency of data science—the case of Data for Good initiatives in the UK nonprofit sector

Data science is a new link in the long chain of quantitative measurement in public policy. The article analyses the promotion of data science as a tool of public policy by operationalizing the instrument constituency framework. The article expands research on instrument constituencies to two new areas: measurement instruments and the nonprofit sector. The empirical analysis focuses on “Data for Good” initiatives in the UK nonprofit sector and is based on 37 interviews with nonprofit data professionals participating I the initiatives. The analysis focuses on actors, instruments, and promises. The findings show that the innovative potential of digital data and data science is central to the initiatives, but the actual practices promoted by the participants are much more varied. The analysis shows blurred boundaries between different promotional coalitions and underscores the collaboration and competition between initiatives. The article confirms that the instrument constituency framework is applicable to the analysis of measurement techniques in the nonprofit sector. It invites further empirical and conceptual work on the unique elements of instrument constituencies that focus on promoting measurement techniques.