Abstract
Recent monetary business cycle models with sticky prices assign no role to monetary aggregates, in the sense that the level of output, price and interest rate can be determined without knowledge of the quantity of money. This paper evaluates the empirical validity of this position by studying the effects of money demand shocks for these variables. We use an identified VAR analysis, isolating money-demand shocks by means of identifying restrictions supported by all models in this class. Contrary to the prediction of the theory, real money balances have substantial predictive power for future movements in inflation and output.