Bureaucratic Discretion Versus Algorithmic Governance: A Comparative Assessment of Rule-Based and Machine-Assisted Decision-Making Frameworks in Public Service Delivery

Abstract

The proliferation of algorithmic decision-support systems within public sector organizations has precipitated a fundamental tension between administrative discretion and computational rule enforcement. This study conducts a comparative institutional analysis of bureaucratic discretion-based frameworks and algorithm-assisted governance models across twelve OECD member states, utilizing a mixed-methods design integrating principal–agent theory, street-level bureaucracy literature, and regulatory accountability benchmarks. Drawing on longitudinal administrative data (2018–2024) and structured interviews with senior civil servants, findings indicate that algorithmic governance enhances procedural consistency and reduces clientelism, yet systematically erodes equity-sensitive judgment in complex casework. Conversely, discretion-intensive regimes demonstrate superior responsiveness to contextual vulnerability but incur accountability deficits. The paper proposes a hybrid adaptive-governance model, delineating conditional domains of appropriate deployment for each framework.