This paper uses a model with utility maximizing households, monopolistically competitive firms, and sticky goods prices to derive a version of Barro and Gordon's time consistency problem for monetary policy where the government's objectives are consistent with a representative household's preferences. The paper applies the methods of Chari and Kehoe to characterize the entire set of time consistent, or sustainable, outcomes. It goes on to find conditions under which the Friedman rule, the optimal policy under commitment, can be supported when the government lacks a commitment technology.