The Missing Middle: Algorithmic Control, Legitimacy, and Worker Resistance in the Nigerian Financial Sector

Abstract

This research challenges techno-utopian discourses around automation by examining experiences, negotiations, and resistance to augmented algorithmic control in the highly regulated, underexamined Nigerian financial sector in the Global South. Employing a qualitative, interpretivist methodology in Lagos, we conducted 49 semi-structured interviews with expert professionals and 8 managers) across two distinct entities, namely a licensed commercial bank and a Bureau de Change (BDC), between February and May 2024. Data were analysed using Braun and Clarke's reflexive thematic analysis. Contrary to assumptions of post-hierarchy, participants regarded dashboards and risk-scoring systems as manifestations of neo-hierarchy, serving as proxy management that is both Taylorist in operation and algorithmic in design, sanctioned by the Central Bank of Nigeria's regulatory intensity and fiduciary requirements. Control failures arose not from excessive monitoring but from insufficient elucidation, resulting in information disparity, a lack of feedback, and preemptive apprehension. The resistance was infrapolitical, as it was neither obvious nor explicit, and took three forms: non-cooperation through analogue manipulation, gaming through data manipulation and use of clients, and negotiation through collaborative algorithmic interpretation. Notably, both frontline staff and managers rejected both full automation and full autonomy, converging on a preference for augmentable control. The theory of this convergence is the missing middle: an explanation is necessary for the relationship contract, including human feedback and contestation learning. We extend the Ouchi (1979) taxonomy to demonstrate how bureaucratic and clan logic exist simultaneously in algorithmic workspaces.