Abstract

The Existence of a World Demand for Money Function. A Reply

This paper comprises an answer to G. Zis’s allegations that our work on the “Existence of a World Demand for Money Function” lacks both theoretical and empirical validity. As shown by the theoretical arguments developed as well as by the econometric tests that were carried out, the allegations of Zis can be completely reversed. We maintain that, whether the behaviour of the world demand for money function is investigated by aggregating a group of countries or by treating these countries in a pooled time series cross-section sample, the testable hypothesis is equivalent. Furthermore, our pooling technique is econometrically superior because it lessens the problem of multicolinearity and distributed lag bias. In fact when an endeavour is made to eliminate part of the distributed lag bias from the G.W.Z. data their results are reversed completely and tend to agree with ours. As to which is the best way of converting national monetary variables to a common denominator, the extensive discussion presented here points to the conclusion that the G.W.Z. approach of using current exchange rates, imparts a positive bias on the income elasticity. Especially on this point we believe that the methodology we followed is better designed to cater for this type of bias.