Answers are grounded in this paper and cite their sources.
Algorithmic and AI systems employed in public administration often manifest technocratic and efficiency premises which may conflict with the democratic and socio-technical fundaments of algorithmic practices in government. Partially, these premises are infused into these public algorithmic systems during the design process. This paper conducts a case study in a consortium of Dutch public organisations that reshapes design practices of algorithmic systems to uncover institutional logics underlying these practices. Our observations and interviews at the consortium of 13 public organisations revealed four design practices that deviate from sociotechnical design practices: (1) distance from politics, (2) underrate the input from non-technical experts, (3) operate on a procurement-oriented dynamic, and (4) follow a chronological phasing combined with agile or scrum practices. Through abductive reasoning, we derived the technocratic and businesslike logics that shaped these design practices. These institutional logics clash with the socio-technical, political, and material nature of public algorithmic systems. To arrive at more responsible, accountable and ethical use of these systems in government, public organisations need to modify their design practices. Our findings provide initial insights into the institutional logics required to change design practices, while also underscoring the complexity of altering routines and practices that are deeply rooted in organisational structures and cultures.
No takes yet. Share an insight, caveat, or question.
Nouws et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: