The Autonomy Budget of Homo Algorithmus: Institutional Conditions of Voter Agency in Algorithmic Democracy

Research on algorithmic democracy has focused extensively on misinformation, polarization, and persuasion (Allcott & Gentzkow, 2017; Sunstein, 2017; Bail et al., 2018). Yet these approaches overlook a fundamental transformation: how algorithmic infrastructures reconfigure the conditions under which citizens form political preferences. The core democratic challenge is not distorted outcomes but eroded voter autonomy. Extending Sartori's (1997) Homo Videns, this paper theorizes Homo Algorithmus - a political subject whose agency is institutionally shaped by algorithmic systems governing visibility, relevance, and choice (Bucher, 2018; Zuboff, 2019). We introduce the "autonomy budget": a finite, institutionally determined resource structuring political agency in algorithmic environments. The framework operationalizes autonomy through four dimensions - transparency (Ananny & Crawford, 2018), user control (Helberger et al., 2018), contestability (Kluttz & Mulligan, 2019), and reflective space (Rosa, 2013; Citton, 2017). Rather than treating autonomy as individual capacity (Christman, 2009; Dworkin, 1988), we specify its institutional preconditions, shifting analysis from voter psychology to the architecture of political information ecosystems (Gillespie, 2014; van Dijck et al., 2018). This framework enables comparative analysis across platforms and regulatory regimes (Gorwa, 2019), treating voter autonomy as a dependent variable shaped by platform governance, algorithmic design, and democratic regulation.