Mill’s harm principle, rationality, and Pareto optimality in games

Abstract Mill’s classic argument for liberty requires that people’s exercise of freedom should be governed by the harm principle (MHP): that is, an action should not harm another. In this paper, we develop the concept of a Millian harm equilibrium (MHE) in n -person games where players maximize utility subject to the constraint of an MHP. Our main result is in the spirit of the fundamental theorems of welfare economics. We show that for every initial ‘reference point’ in a game the associated MHE is Pareto efficient and, conversely, every Pareto efficient point can be supported as an MHE for some initial reference point. This is an important result for an old question in political philosophy over whether the exercise of liberty is consistent with order in society and for how we think about policy in a non-ideal world.