It is generally accepted that agency requires materiality, since action must originate somewhere. In group agents, this implies that they are agential material objects rather than hovering, mind-like entities (Hess 2025). I argue that the material existence of group agents can be best explained by a Baker-style constitution relation (2000) between a group and its members, which is strictly non-mereological. First, I argue against Collins (2023) that her mereological account of group agents cannot successfully explain the relation between a group and its members because it fails to meet the non-transitivity desiderata of group agents. I subsequently show how this critique can be generalized to all mereological proposals (Hawley 2017; Hansson Wahlberg 2014). Secondly, I argue that the existing constitution accounts of group agents either fail to identify the correct material object from which the action of a group originates (Epstein 2015; Hindriks 2013; Uzquiano 2004) or account for it through mereological views of constitution (Harris 2020), which run into the same aforementioned problem. I conclude that a strictly non-mereological constitution account of the material existence of group agents is the most explanatorily successful since it is the only proposal that can meet the non-transitivity desideratum of group agents.
