Abstract
This article investigates the relationship between sum, Divisia, and monetary velocity money, prices, and income using the notion of Granger (1969) causality. This is achieved by evaluating empirically (using Dickey–Fuller unit-root tests) whether the macroeconomic time series under consideration are trend stationary or difference stationary and by conducting tests using three different, ad hoc, lag lengths—8, 6, and 4 quarters—as well as a statistically determined—using Akaike's (1969a,b) final prediction-error criterion—lag structure.