The Role of Causal Features in Strategic Classification for Robustness and Alignment

Sophia Günlük, António Góis, Nir Rosenfeld, Nidhi Hegde, Simon Lacoste-Julien, Dhanya Sridhar

TL;DR: We identify conditions where a causal classifier gives optimal post-adaptation loss for a range of adaptations, we provide a decomposition of cross-entropy risk, and we study alignment between the predictor and predicted users.

Abstract: In strategic classification, an institution (e.g., a bank) anticipates adaptation from users who change their features to increase utility in a classification task (e.g., loan repayment). Since a key challenge is the distribution shift induced by users, we turn to causal models, which have been shown to bound the worst-case out-of-distribution (OOD) risk, and establish several new results that link causality and strategic classification. First, we show that causal classification leads to optimal classification error after any sufficiently large adaptation, when the noise is bounded in a certain way. Second, when these assumptions do not hold, we show OOD cross-entropy risk of optimal classifiers decomposes into an OOD bias term and a term arising from not using all observable features, allowing us to determine when causal classifiers have an advantage. Finally, we show that causal classifiers can align long-term incentives between institutions and users, contrasting with previous work that highlights social costs of such approaches. We validate our theory empirically on synthetic data, finding that our results predict behavior in practice.

Submission Number: 1029