Algorithmic Fairness in Performative Policy Learning: Escaping the Impossibility of Group Fairness