A Stable Velocity Function for Canada? A Further Note

Abstract

Dr. Goodhart's computations clearly indicate that the velocity function (not income velocity itself) is shifted (becomes unstable?) when the institutional framework is changed in a significant way, as appears to have happened in Canada as a result of the implementation of important sections of the Report of the Porter Commission. Interestingly, there appears to have been a shift of the same kind around 1935 when some of the recommendations of the Report of the Macmillan Commission of that same year were implemented. I suspected something like this when I unsuccessfully tried to predict (or post-dict) the income velocity of money before 1935 with the velocity function I had estimated using the data for the period 1935-59. If I am roughly correct in my supposition of a simple shift in the velocity function, one could still consider that the function is stable, since a dummy variable would account for the effects of Royal Commissions and of other significant institutional changes. It is true, however, that even though such a variable would rescue the equation for some purposes, as a guide to policy much of its usefulness would be destroyed, since the exact quantitative impact of significant institutional changes would have to be re-estimated with every new Royal Commission. On the other hand, it may be that the shift in the function could be greatly reduced by using a narrower definition of the money supply, namely one that was restricted to currency and demand deposits; this change, and the introduction of the yield on saving deposits as an explanatory variable in the velocity function, could rescue it for policy and for other purposes. The problem here is that the statistics on demand and saving deposits in Canada have been published separately for recent years only.