Substitutability, equilibrium transport, and matching models
Alfred Galichon, Antoine Jacquet
TL;DR
The chapter investigates how substitutability shapes equilibrium computation in economic models, highlighting that coordinate-update methods like Jacobi's algorithm and Sinkhorn's algorithm converge under Z- and M-function structures. It develops transferable-utility matching as a fixed-point problem using a distance-to-frontier transform to eliminate transfers, enabling Jacobi-based solutions with convergence guarantees. For non-transferable utility, it ties stable matching theory to monotone operator frameworks via Adachi's formulation and DALM, establishing fixed-point characterizations and convergence results for equilibrium matchings. The text offers both theoretical convergence results and practical algorithms (Sinkhorn, Gale–Shapley, DALM, Adachi) and illustrates them with applications to expenses, housingMarket rent control, and dynamic ride-hailing pricing. Collectively, it provides a unified, mathematically rigorous toolkit for analyzing substitutability-driven matching under both transferable and non-transferable utilities.
Abstract
This chapter explores the role of substitutability in economic models, particularly in the context of optimal transport and matching models. In equilibrium models with substitutability, market-clearing prices can often be recovered using coordinate update methods such as Jacobi's algorithm. We provide a detailed mathematical analysis of models with substitutability through the lens of Z- and M-functions, in particular regarding their role in ensuring the convergence of Jacobi's algorithm. The chapter proceeds by studying matching models using substitutability, first focusing on models with (imperfectly) transferable utility, and then on models with non-transferable utility. In both cases, the text reviews theoretical implications as well as computational approaches (Sinkhorn, Gale--Shapley), and highlights a practical economic application.
