A comparison of the rational expectations and the general-to-specific approaches to modelling the demand for M1

The issues involved in assessing the relative performance of forward-looking (rational expectations) models and backward-looking error correction models based on the general-to-specific methodology are examined. Using the demand for M1 in the United Kingdom as a specific example, the authors find that both the error correction model and the forward model perform adequately, although a version of the error correction model does encompass the forward model in a nonnested test. The authors discuss reasons why the discrimination between the two models may prove difficult in practical situations using finite data sets. Copyright 1992 by Blackwell Publishers Ltd and The Victoria University of Mancheste