An error-correction model is used to study the long- and short-run determinants of U.S. demand for M2. The money demand function presented here exhibits parameter stability and predicts quite well the actual behavior of M2 growth in the 1980s.

Granger-causality tests used here find that: [1] unit labor costs add no predictive power to inflation forecasts; and [2] the gap between actual and potential output does help predict inflation, but only in the short run.

Concerns that interest rates are too high have been prevalent throughout the 1980s. Even after adjusting for expected inflation, many people argue that real interest rates are inordinately high by historical standards. Yash Mehra, in his article “The Tax Effect and the Recent Behaviour of the...

Conventional M1 demand functions reformulated using error-correction and cointegration techniques neither depict parameter stability nor satisfactorily explain short-run changes in M1. Thus, M1 remains unreliable as an indicator variable for monetary policy.