1: \begin{abstract}

2: %

3: Neglecting the effect that decisions have on individuals (and thus, on the underlying data distribution) when designing algorithmic decision-making policies may increase inequalities and unfairness in the long term---even if fairness considerations were taken in the policy design process.

4: %

5: In this paper, we propose a novel 

6: framework for achieving long-term group fairness in dynamical systems, in which current decisions may affect an individual's features in the next step, and thus, future decisions. 

7: %

8: Specifically, our framework allows us to identify a time-independent policy that converges, if deployed, to the \emph{targeted} fair stationary state of the system in the long-term, independently of the initial data distribution.

9: %

10: We model the system dynamics with a time-homogeneous Markov chain and optimize the policy leveraging the Markov chain convergence theorem to ensure unique convergence.

11: %

12: We provide examples of different targeted fair states of the system, encompassing a range of long-term goals for society and policy makers.

13: %

14: Furthermore, we show how our approach facilitates the evaluation of different long-term targets by examining their impact on the group-conditional population distribution in the long term and how it evolves until convergence.

15: %

16: \end{abstract}

17: