A 1992 paper by Yoshihisa Baba, David Hendry, and Ross Starr presents a model of the demand for narrowly defined money (Ml) in the United States that shows a dramatic improvement in both fit and stability over earlier models. This note estimates an alternative model that is based on the same data set, uses a similar error-correction methodology, and has very similar statistical properties to the original. Both models show remarkable stability throughout the past three decades.
Two conclusions are that the improvements are due more to the use of complex dynamics than to the introduction of variables representing financial innovation (as alleged by Baba, Hendry, and Starr) and that some of the economic properties are not robust with respect to minor changes in specification. Notably, whereas in the original model, money demand has a low elasticity to real income, which suggests substantial economies of scale in money holdings, this alternative has a unitary elasticity, which suggests that money and income should grow proportionately in the long run.