The Hahn problem is that, even if a “monetary” equilibrium with valued fiat money exists in general equilibrium, a “nonmonetary” equilibrium with a zero price on money generally also exists; why should we expect the former over the latter? Here, I consider the preferences that will survive repeated trading in an exchange economy where agents compete in biological fitness. With unobservable preferences and positive assortativity in matching, evolutionarily stable preferences implement the competitive equilibrium that maximizes the sum of agents' fitnesses. In a standard Bewley–Townsend model, this implies selection of the monetary over the nonmonetary equilibrium, and also implies the survival of agents with “money in the utility function”. Trust in fiat money is only a recent development, and even today such faith is hardly universal. ( Cass and Shell, 1980 )