The long-run and short-run demand for money: Additional evidence

This paper presents estimates of Chow's money-demand equations using Chow's data. The equations are adjusted for autocorrelation using two autocorrelation transformations, the standard Cochrane-Orcutt transformation that deletes “initial” observations, and one that does not delete these observations. The estimates focus on the question of the asset versus the transactions specifications. The results reaffirm Chow's original conclusion which supported the asset motive and, thereby, reversed nearly all of the findings recently reported by Lieberman (1980). The paper concludes that one should be wary about using the Cochrane-Orcutt transformation, especially when the ratio of the number of deleted observations to total observations is large.