Designing for diversity: team formation and participatory policy design

Participatory policy design benefits from tools that engage diverse voices. one emerging tool is the civic hackathon, which brings together diverse participants to co-develop policy ideas. For such events to succeed, teams need to be balanced in skill and size while also respecting participants' preferences. We compare two civic hackathons: one with self-assigned teams and another with algorithmic team assignment. Self-assignment produced large imbalances in size and skill; algorithmic assignment created dissatisfaction and led to chaotic self-reassignment. Drawing on our findings and the broader literature on teams and team-formation algorithms, we propose a two-stage assignment process combining algorithmic matching with a structured opportunity for swaps. We outline a research agenda to test this approach in future civic hackathons. this agenda offers value not only to those interested in participatory policy design, but also to scholars examining team diversity, hackathons, and algorithmic team assignment.