["Nations and Nationalism, Volume 32, Issue 2, Page 374-385, April 2026. ", "\nABSTRACT\nWhy does power‐sharing lead to peace and effective governance in some cases but not others? Whereas the current literature on this question predominantly focuses on institutional design, this article argues that more attention should be given to the everyday activities, routines and processes through which power‐sharing is operated. Defining power‐sharing as a way of governing that combines inclusivity and joint decision‐making, the article argues that variations in the performance of power‐sharing result from the work deployed by various actors to achieve or resist elite cooperation across conflict lines. Three practices are essential in determining elites' ability to cooperate: the (convergent or divergent) socialization of power‐holders, the (polarizing or unifying) representation of their supporters and the (constructive or obstructive) interpretation of power‐sharing rules. This argument is evaluated empirically against two cases of post‐conflict, ethnic power‐sharing presenting variations in performance that cannot be explained by institutional design alone—Burundi and Northern Ireland.\n"]