CTESS Seminar

Abstract: We introduce a novel perspective by linking ordered probabilistic choice to copula theory, a mathematical framework for modeling dependencies in multivariate distributions. Each representation of ordered probabilistic choice behavior can be associated with a copula, the analysis of representations through established results from copula theory. The connection highlights extremal representations–associated with the Fréchet-Hoeffding bounds–and their distinctive structural properties. Thus, we derive identification methods to uniquely determine the specific heterogeneous choice types and their corresponding weights. The unified framework elucidates the uniqueness of known representations while showcasing the potential of copula-based methods to uncover new choice models and results. Our analysis provides tools for inferring micro-level behavioral heterogeneity from macro-level observable data.

Written with Christopher P. Chambers and Kemal Yildiz.

This seminar is sponsored by the Center for Theoretical and Experimental Social Sciences (CTESS), which is part of The Ronald and Maxine Linde Institute of Economic and Management Sciences.