How Does Ideology Impact the Design of Algorithmic Governance?

Abstract

Algorithmic management and algorithmic governance have emerged as prominent topics of concern over the past decade. The notions of algorithmic management and algorithmic governance are particularly salient in the age of platform ecosystems and flatter hierarchies enabled by IT’s affordances (Turco, 2016). Public attention algorithmic management often focuses on concerns of programmers and designers “encoding their biases and assumptions” into the design of social-technical systems. This research seeks to uncover how designers’ assumptions and ideologies impact the design of algorithmic governance systems. Prior research predicts that top-down “managerial” logics would be encoded into the design, resulting in feuding or “loose coupling” by end-users who refuse to be controlled (Berente & Yoo, 2012). However, in my empirical setting, I identify the emergence of a “bottom-up” governance design that was heavily influenced by multiple logics and points of view. While prior research has tended to black-box the impact of “ideological work” on the design and implementation of information systems, my data speaks strongly to this being a central element of the design process. Drawing on a six-month ethnographic study of a community of blockchain developers, I ask: what roles do ideology and worldview play in the design and implementation of algorithmic governance systems?