Group-Fair Classification with Strategic Agents

The use of algorithmic decision making systems in domains which impact the financial, social, and political well-being of people has created a demand for these to be "fair"under some accepted notion of equity. This demand has in turn inspired a large body of work focused on the development of fair learning algorithms which are then used in lieu of their conventional counterparts. Most analysis of such fair algorithms proceeds from the assumption that the people affected by the algorithmic decisions are represented as immutable feature vectors. However, strategic agents may possess both the ability and the incentive to manipulate this observed feature vector in order to attain a more favorable outcome. We explore the impact that strategic agent behavior can have on group-fair classification. We find that in many settings strategic behavior can lead to fairness reversal, with a conventional classifier exhibiting higher fairness than a classifier trained to satisfy group fairness. Further, we show that fairness reversal occurs as a result of a group-fair classifier becoming more selective, achieving fairness largely by excluding individuals from the advantaged group. In contrast, if group fairness is achieved by the classifier becoming more inclusive, fairness reversal does not occur.