A reconsideration of Sims' evidence concerning monetarism