Digital Participation, Deliberation and Legitimacy: A Theoretical Account of Civic Technology, Algorithmic Mediation and Citizen Voice in Public Decision-Making

Abstract

Governments and civil-society organisations increasingly use digital platforms to consult citizens, crowdsource priorities, report service failures and deliberate on budgets, and artificial intelligence is now being added to summarise submissions, cluster opinions and moderate discussion. Enthusiasm has outrun theory. E-participation research treats participation mainly as an input channel, deliberative theory says little about public administration, and work on AI-assisted deliberation is split between promise and rejection, so no account explains what digital and algorithmic mediation do to participation and to the legitimacy participation is meant to produce. This article supplies one. It uses a conceptual, theory-building design based on an integrative review of 33 peer-reviewed journal articles published between 2023 and 2026 and one primary statute, combined with typology construction, proposition derivation and an illustrative application to African local government. Drawing on deliberative-democratic theory, participatory-governance scholarship and recent research on civic technology, it distinguishes three functions of participation: informing decision-makers, legitimating decisions and developing citizens. It argues that digital participation serves the first efficiently, the second ambiguously and the third poorly, because reciprocity, reflection and the obligation to justify are precisely what platform and algorithmic mediation economise on. Seven propositions specify when digital participation enhances rather than simulates legitimacy, how algorithmic summarisation and opinion clustering shift whose voice is heard, and how digital channels interact with ward committees, town halls, traditional councils and participatory budgeting. Applied to South Africa, Kenya, Nigeria and Ghana, where participatory mandates are expanding while digital access remains uneven, the account locates the decisive weakness in the link between forums and decisions rather than in the channel. The article recommends published justification registers, hybrid sequencing of digital and synchronous deliberation, contestable algorithmic mediation secured through procurement, and formal recognition of ward-level and traditional intermediaries, and it sets out how each proposition can be tested.