Answers are grounded in this paper and cite their sources.
As algorithmic decision-making (ADM) becomes increasingly embedded in public administration, expectations of efficiency and predictive accuracy have grown alongside concerns about fairness, accountability, and legitimacy. Existing scholarship, however, has tended to treat these tensions as discrete problems rather than as interlocking dynamics that mutually reinforce one another. Drawing on Stone’s policy paradox as an analytic lens, this article conducts a single revelatory case study of South Korea’s e-Child Happiness Support (e-CHS) system. Through qualitative analysis of government documents, a thematic analysis of online parental discourse, an expert interview, and triangulated secondary qualitative evidence on frontline workers, the authors identify three paradoxes and, critically, the feedback loop that binds them. The technology–humanity paradox emerges from design flaws and data proxies that produce algorithmic accuracy without legal or contextual validity. The policymaker–target group paradox arises when preemptive protection is experienced as intrusive surveillance,particularly because welfare-eligibility data and abuse-detection data structurally overlap,reproducing an “automating inequality” effect. The policymaker–implementer paradox surfaces in the gap between expected efficiency and the heightened accountability borne by under-resourced frontline officials. Based on these findings, the authors outline the theoretical and practical implications of ADM adoption in the public sector.
No takes yet. Share an insight, caveat, or question.
Eom et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: