Abstract Much has been written on how an active central bank produces inflation outcomes above and beyond what commitment policy would produce. This paper contributes to this body of literature by simulating from the state estimates of both commitment and discretionary policy equilibria in a familiar dynamic New–Keynesian framework. Optimal interest rate and inflation rate policies are derived under the two regimes for six developed economies. The model is estimated using Bayesian methods employing a random-walk Metropolis–Hastings algorithm. Optimal inflation and interest rate policies for each of the economies are simulated. Results suggest that the simulated inflation induced by discretionary policy is not significantly different from commitment policy after 2000 for five of the six countries (including the U.S). Simulated commitment interest rate policy is on average 1.9% higher at the center of the distribution, suggesting that discretionary interest rate policy is on average more often loose compared to commitment interest rate policy. Simulations of the average inflation deviation and welfare loss of discretion policy indicate are greatest when the central bank exhibits low preference for inflation targeting and high preference for output stability.