New Fisher Market Model Guarantees Fair Labor Equilibrium with Fast Algorithms

A team of computer scientists from Columbia University and the University of Illinois at Urbana-Champaign has developed a novel theoretical framework that models two-sided labor markets as a hybrid of goods and chores Fisher markets. In their paper, “Competitive Equilibrium in Labor Economies through the Lens of Goods and Chores Fisher Markets,” the authors treat each task as a good for the user (who derives utility from it) and a chore for the worker (who incurs disutility to perform it for earnings). This dual nature allows endogenous pricing that balances demand and supply, guaranteeing the existence of a competitive equilibrium (CE) in very general settings. The paper also proves that both the first and second welfare theorems hold, meaning any CE is Pareto optimal and any Pareto optimum can be supported by some prices.

For the computationally important case of linear preferences, the authors initially faced a non-convex optimization problem—similar to the known difficulties in chores-only markets. Remarkably, they discovered a set of surprisingly positive results. First, they designed a polynomial-time combinatorial algorithm that updates prices using a Walrasian tâtonnement scheme and, under equal-income (CEEI-like) assumptions, runs in strongly polynomial time. Second, they showed that the non-convex labor-market program admits a change of variables that transforms it into a linear program (LP). Although the LP has irrational coefficients, the team provides an efficient method to handle them. This is notable because even for goods-only linear Fisher markets, obtaining such an LP formulation remains an open problem. The work not only advances algorithmic game theory but also offers practical tools for designing fair and efficient digital labor platforms and task markets.