Designing Long-term Group Fair Policies in Dynamical Systems
This addresses fairness degradation in sequential decision systems for policymakers and society, representing a novel methodological contribution rather than an incremental improvement.
The paper tackles the problem of algorithmic decision-making policies increasing inequalities over time by proposing a framework for achieving long-term group fairness in dynamical systems where decisions affect future individual features. The result is a time-independent policy that converges to targeted fair stationary states regardless of initial data distribution, with examples showing how to evaluate different long-term societal goals.
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. In this paper, we propose a novel 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. Specifically, our framework allows us to identify a time-independent policy that converges, if deployed, to the targeted fair stationary state of the system in the long term, independently of the initial data distribution. 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. We provide examples of different targeted fair states of the system, encompassing a range of long-term goals for society and policymakers. 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.