To be useful as policy indicators, the monetary aggregates M1 and M2 must either (1) provide information on the current state of the economy or (2) help predict future values of economic variables that are of concern to the policymakers. Statistical analysis shows that these aggregates perform poorly in the first task but somewhat better in the second. They work poorly as information providers because the underlying money demand functions upon which their performance depends lack the required statistical properties of temporal stability and precision-in-estimation. They work better, however, as forecasters of future economic activity, M2 more so than M1.
Amanda L. Kramer
Order single copies or subscribe to Economic Quarterly and other publications from the Federal Reserve System.