Between Algorithms and Autonomy: A Sociological Comparison of the Experience of Freelancers from Ghana and South Africa on the Upwork Platform.

Platform work is an increasingly important livelihood across Africa. Platforms like Upwork offer the possibility of working when, where, and how workers choose, but this promise of autonomy is complicated by opaque scoring systems, client reviews, and platform visibility. Drawing on semi-structured interviews with eighteen Ghanaian and South African knowledge workers, this thesis analyses how freelancers on Upwork understand and make sense of the apparent trade-off between worker autonomy and platform control. Participants’ narratives were treated as interpretations of experience rather than factual reports of events. The findings show that the promised freedom of platforms is often constrained by opaque scoring mechanisms, client control, unpaid visibility labour, and economic dependence on the platform. Ghanaian respondents often described Upwork as a response to economic pressure and weak institutional protection, while South African respondents more often framed it in terms of career opportunities and regulatory mismatch. Stress around the Job Success Score and fear of negative reviews were central issues in both national contexts. The workers described here do not passively try to overthrow platform control; rather, they adapt their work practices in response to platform logic to maintain earnings and visibility. Using labour control theory, the thesis demonstrates that control on Upwork relies less on direct supervision and more on platform rules, client demands, and freelancers’ self-adjustment.